Advancing Risk Communication with Decision-Makers for Extreme Tropical Cyclones and Other Atypical Climate Events: Proceedings of a Workshop (2024)
Chapter: 6 Practical Translation of Risk in the Public Arena
Chapter 6
Practical Translation of Risk in the Public Arena
“How do we determine how to best bridge research and practice to advance risk communication?” was the first question posed by Gina Eosco, Division Chief of the Science, Technology, and Society Division, National Oceanic and Atmospheric Administration (NOAA), during her presentation on public initiatives to evaluate the effectiveness of risk communication products. This question encapsulated the theme of the second day’s first session: Practical Translation of Risk in the Public Arena. Panels on the work of translating risk from research to practical messaging and responses included discussions of risk communication innovations and new frontiers in communication around tropical cyclones in the public and private sectors, as well as descriptions of specific communications technologies developed by private companies.
RISK COMMUNICATION INNOVATIONS AND NEW FRONTIERS IN TROPICAL CYCLONE COMMUNICATION: PUBLIC SECTOR
The efficacy of technologies and approaches that facilitate communication of risk—and uncertainty—to the public would likely benefit from being evaluated. This panel, composed of representatives from the National Hurricane Center (NHC) and NOAA, addressed this topic from the public-sector perspective.
Andrea Brennan, Director, NHC, began the panel by discussing the work of communicating risk at the national level in the event of a tropical cyclone. The NHC tracks the entire “lifecycle” of the event, which means that risk communication often spans multiple hazards that vary in severity across time. This work requires the coordination of simultaneous different messages to multiple populations in multiple locations, because experiences of stages and impacts differ at any given point in time and by location.
The timing of risk communication is an important factor in helping individuals and communities take protective action. Risk communication faces extra challenges around “low-probability, high-consequence events,” Brennan noted. Certain-
ty about where and when extreme impacts will occur is usually available only a few hours before those impacts begin— “well beyond an actionable timeframe for preparations.” At 36-48 hours out, for example, a great deal of uncertainty is present, and so risk is communicated across a large geographical area even though the threat (e.g., high winds, storm surge) will ultimately only affect a small area.
Therefore, the tools used by the NHC to convey risk range from early forecasts to in-the-moment messaging. Brennan mentioned four tools that cover a range in terms of time and certainty. First, the 7-day tropical weather outlook is a probabilistic forecast that serves as an early alert that storm systems are forming. Messages are crafted to raise awareness of the storm formation, define a broad area of potential impact, and emphasize preparedness. The National Weather Service (NWS) coordinates the messages, with the aim of providing consistent information for use by the broadcast meteorology community, emergency managers (EMs) at the state and local levels, and other media.
The official storm forecast—which captures the forecast of the track, the intensity, and the size of the storm, as well as the cone that indicates the possible track of the center of the storm, is another such tool. This tool is also deterministic, she noted, and does not incorporate uncertainty.
A third category of tool is the “variety of probabilistic hazard-based products” that focus on specific, individual hazards associated with tropical cyclones—including storm surge, wind, flooding, rainfall, and tornadoes. These products do incorporate information about uncertainty, and they focus on the hazards present rather than predictions of the overall storm’s track. At 3 to 5 days out, messaging becomes more focused on areas where the highest impact is anticipated; however, even at this point, “it’s still too uncertain to get very specific about exact timing and magnitude.” These messages also indicate how the risk is changing as the storm develops, she explained. This 3- to 5-day timeframe is often when decisions about preparations are made, and individuals and communities begin to take protective actions—especially when, for example, evacuation plans require a long lead time. At this point, Brennan noted, the NHC’s messaging emphasizes the importance of following the advice of local officials and uses “rather severe wording” if needed, for example, when confidence is high that the event will be life-threatening.
Watches and warnings comprise the fourth type of tool used to communicate risk—in this case, about the risk of a particular hazard at a specific location. These messages become more detailed and vary depending on the time and location they cover. The focus here is more on the hazards than on the track or intensity of the overall storm, said Brennan. Each hazard requires a different response, and likely each location will be impacted at a slightly different time. She described how, during Hurricane Harvey, messaging from the NHC had to cover both the risk to the mid-Texas coasts of extreme wind and storm surge at landfall, and, several days later, the risk to the upper Texas coast of catastrophic flooding. Lead time is critical, and Brennan noted that the NHC uses watches and warnings in an effort to officially provide at least 36-48 hours of lead time. She echoed comments from previous discussants, including Pearson, Strickland, and LaDue, that
communities often must begin to take protective action, such as evacuation, well before this mark.
Post-storm messaging is also critical, Brennan noted. This messaging might cover cleanup safety, heat risks, post-storm generator safety, and other topics. Raising these topics before the storm hits is important because reaching people once they are affected can be very difficult.
Alongside forecasts, messaging tools in and of themselves are critical to communicating risk in the event of a tropical cyclone. Brennan referred to “discussion products,” which are designed to help forecasters convey information about uncertainty and communicate “how the risk is changing as the storm evolves.” The NHC also provides impact-based decision support briefings to federal and state agencies, including the Federal Emergency Management Agency (FEMA), and sometimes local officials. Social media channels, including livestreams, are also an important part of the communication strategy and often are active in advance of the NHC’s media pool, which is more formal and often “more focused on the last couple of days before landfall.”
Castle Williamsberg, the second speaker on the panel, discussed challenges to “translating risk communication research for practitioner use,” particularly with the aim of using research in the social, behavioral, and economic sciences to modernize and improve the Tropical Cyclone Product Suite at the NHC.1 He described work by the Weather Program Offices (WPO) Social Science Program at NOAA, which funded four complementary studies, designed with intentional overlaps and differences, to yield a body of social science research on the impact of risk messaging products.2 WPO and NWS social scientists synthesized the results of these four studies and worked with the NWS Tropical Roadmap Team to ensure that findings and future research were conceptualized in ways that were “operationally relevant and could be used in practice.” Practitioner partners include EMs, forecasters, local officials, and other decision-makers.
Williamsberg highlighted five of the key themes from these studies as framed in operationally relevant terms. First, probabilistic information helps people make decisions in the midst of uncertainty. Practitioners could briefly explain how to interpret probability information, rather than simplifying their message and leaving out uncertainty all together. Second, partners expressed a strong desire for more localized information, even when level of uncertainty about the local forecast is high. Third, different types of “timing information” were critically useful to partners making decisions about risk communication. This information included not only when hazards might begin or reach the highest impact, but also when they might end and their duration (understand the impacts of sustained elements—for example, the impact of sustained wind to bridge infrastructure). Fourth, in situations of more than 5 days of lead time, partners wanted more information about
___________________
1 More information about Tropical Cyclone Products at the NHC is available at https://www.nhc.noaa.gov/productexamples/ and https://www.nhc.noaa.gov/cyclones/.
2 More information about the Weather Program Offices Social Science Program at NOAA is available at https://wpo.noaa.gov/social-science/.
the forecast models and forecast confidence. Linked to this were changes, suggested by partners, to NOAA’s Tropical Cyclone Product Suite that would help “optimize the extraction of key information from textual and graphical products to further encourage message transmission.” Finally, partners placed a high value on summary products— “key message products.” A single dashboard or landing page that brought together “the entire ecosystem of tropical products” could help with message transmission.
Williamsberg then highlighted four challenges to translating research into actionable items for practitioner use that emerged from the synthesis of the studies’ findings. The first challenge is understanding when researchers have enough information to ensure that study findings are usable by the practitioners. The second challenge involves addressing operational challenges and ensuring that the findings are “operationally relevant” (i.e., findings that are practicable by EMs, local officials, forecasters, and other partners). In collaboration with the NWS’s Tropical Roadmap Team and with meteorologists, the WPO group developed findings that were beneficial to those practitioner partners and to “help accelerate their research-to-practice process.” The third challenge relates to how best to share the findings among the intended audience, including meteorologists, EMs, local officials, and other decision-makers. The WPO also developed the System for Public Access to Research Knowledge (SPARK).3 In addition, Williamsberg said, the WPO is developing a story map that will help make findings available to practitioners so that they, in turn, can be more attuned to what their own audiences and partners know or are focused on. The last challenge identified is how best to track the successful transfer of knowledge and the benefits associated with it. This effort would involve studying how practitioners incorporate research information into their work, learning how best to provide practitioners with research findings in usable forms, and tracking their usage of this new knowledge (see Porter et al., 2024).
Gina Eosco, also with NOAA’s WPO, addressed the question of how to evaluate the effectiveness of risk communication. Risk communication is “inherently multi-sector,” she noted, and therefore, to be effective, requires leveraging partnerships across the academic, public, and private sectors both in delivering messaging and in determining what kind of an impact such messaging makes. The first step, she noted, is establishing how success is measured: in this case, by looking at the impact of messaging during past events. To this end, WPO is developing the Societal Data Insights Initiative (SDII), a social science data infrastructure that integrates and synthesizes social and meteorological data, which enables users to gain insights into the societal effects of various products and services. Eosco emphasized that this endeavor is collaborative.
One challenge to the work of researching societal impact is gathering large, generalizable samples that more closely represent the public audience being served by these products. Eosco highlighted the importance of longitudinal research and event-based research, such as the studies outlined in Chapter 2. Partnering with
___________________
3 More information about the SPARK platform is available at https://wpo.noaa.gov/empowering-open-science-unveiling-wpos-system-for-public-access-to-research-knowledge-spark/.
private-sector companies that develop weather apps to embed a survey in their app might, she suggested, yield a broader and more representative sample.
An important topic of research is how the public acts, and when, in response to specific products, Eosco said. Data on this topic are not available; “we need more evaluation capacity, and we need varying types of data to help answer these questions.” In this, too, she suggested, a collaboration among public, private, and academic partners—who could share data (e.g., commercial, financial, retail)—that might yield better insights into how people respond to risk communication and where gaps might exist.
Another challenge is present in the idea that “the publics and partners make better-informed decisions when we communicate uncertainty,” a theme raised repeatedly in earlier sessions. Eosco noted a fundamental disagreement: research shows that communicating uncertainty and probabilistic information leads to the best decision-making practices on the part of the public and partners. “Respected partners,” on the other hand, consistently advocate for a simple message and deem probabilistic information either not useful or confusing. A fundamental disagreement around the effectiveness of risk communication, and how best to understand it, occurs if research says one thing and partners do not believe it, Eosco said. She stressed that the resolution to this question does not lie in determining which side is right or wrong; she notes there are likely limitations on both sides: bias and sample challenges on the research side, and heuristics that influence understanding on the public side. The more productive question, Eosco argued, is how the WPO might better understand the gap and why it exists, and how to “find that space of understanding.” Trusted relationships between the public and private sector already support this approach, she noted. Eosco ended her presentation by explaining that NOAA and NWS are undertaking a paradigmatic shift in communication models, from a “one-size-fits-all” approach to a more personalized or localized model that addresses specific and unique needs. This effort requires knowledge of what those unique needs are, she explained, which happens over time and through strong partnerships.
Communicating uncertainty and probabilistic information to partners was the focus of the question-and-answer session, which centered on a query from Robby Goldman, an audience member, who wondered about examples of such communication done simply or in actionable ways. Eosco responded that all products are based on probabilistic information, even if that is not explicit in the outward-facing message, and that there are many ways to communicate uncertainty effectively. She noted that several messages and products used by the NHC incorporate uncertainty already. She pointed to Communication Assist Techs—CATs—that help people gain a better sense of whether their area will, for example, experience a dangerous storm surge.
Brennan added that products vary on their explicitness about the probabilistic information that underpin all of them; when uncertainty is not foregrounded, this is often an effort to translate risk into “a clearer, actionable message.” Products that make probabilistic information explicit are, she noted, sometimes intended
primarily for use by EMs and other experts who may be more interested in the specific details as they make decisions. “The goal is to put all of that probabilistic information out there for people to exploit, but not necessarily to put it all in the public-facing products,” Brennan concluded.
MESSAGING TECHNOLOGY WALKTHROUGH
The second panel of the session involved demonstrations of new and emerging technologies in the risk communication field. Mike Gerber, Wireless Emergency Alert Expert and Meteorologist, NWS, the first speaker, described the many steps involved in distributing Wireless Emergency Alerts (WEAs) as used in the NWS Office of Dissemination and other alerting entities. WEAs are activated only during hazards that pose a great threat to life and property. In the tropical storm category, these include warnings for hurricanes and typhoons, as well as extreme wind. WEAs are template-based, with specific details filled in. There are both English and Spanish templates, as well as shorter and longer message lengths (90-character and 360-character templates). These messages are received by almost every cell phone.
WEA alerts involve cross-sector collaboration. The alerting process begins with a message originated by official alerting authorities, including more than 1,600 federal authorities, state agencies, territorial agencies, tribal governments, and local authorities. Using third-party authoring software, officials convert the message to common alerting protocol (CAP) format, which is based in XML and serves as the international standard format among alerting technologies. This CAP message is then sent to the Integrated Public Alert and Warning System (IPAWS), a CAP message aggregator run by the Federal Emergency Management Agency (FEMA).4 At IPAWS, the CAP message is converted to another format—Commercial Mobile Alert for C-Interface (CMAC), which is specific to wireless alerts. The message in the CMAC format is attached to a polygon that indicates the geographical alert area. From here, the message moves from public agencies to private corporations within the wireless industry, who map the alert polygon onto their own network topology to identify appropriate cell towers from which to broadcast to the WEA. Phones receiving this message do not automatically display it; instead, they use device-based geofencing (DBGF), a process by which the phone compares its location, if known, with the polygon, to determine whether or not the user is actually in or close to the targeted area. If the user is inside this area, the phone displays the alert. This cuts down on over-messaging, as when, for example, the cell towers reach a larger area than the polygon. Gerber concluded by noting that, also in an effort to avoid over-messaging and desensitization to alerts, the NWS is focusing on streamlining messaging and using an impact-based warning approach to target affected areas.
Brock Aun, Vice President of Communication and Public Policy, HAAS Alert, then presented on Safety Cloud, a digital alerting delivery system that delivers re-
___________________
4 More information about the FEMA IPAWS can be found at www.FEMA.gov/IPAWS or by inquiry to IPAWS@fema.dhs.gov.
al-time messages to drivers, built by HAAS Alert.5 HAAS Alert has partnered with navigation apps WAZE and Apple Maps, which incorporate Safety Cloud into their own platforms and alert drivers to crashes, emergency vehicles, work zones, and other road hazards. Delivering alerts in this way gives drivers a small bit of advance warning, which, Aun said, can drastically reduce the odds of a crash in response to these various hazards. Aun pointed to a 2021 study by researchers at Purdue University that showed that, compared with the light bars and sirens found on most emergency vehicles, the digital warning helped reduce hard braking next to a roadside incident by 80 percent. The study estimated that only about 30 percent of drivers received the alerts, but even this minority was enough to influence behavior of most drivers. This particular message delivery platform offers several technical benefits, Aun noted: it aggregates information for automakers, it integrates into already-existing cloud code, it is a cross-platform solution to message delivery, and it has been integrated with IPAWS in certain areas to translate IPAWS data into real-time messaging. Safety Cloud is “aggregating the universe of roadway hazards and providing a single, authoritative stream for automakers.” This effort involves building partnerships with automakers, who can incorporate this single system into various other connective systems already in place within the car. Safety Cloud integrates with more than 50 platforms already in vehicles on the road, including fire trucks, thanks direct work with emergency vehicle manufacturers to integrate Safety Cloud into platforms already in use. Aun noted that, throughout this effort, the main goal is not to provide every driver with a single platform, but rather, to connect the multitude of platforms already in use, making an “interoperable network of safety.”
Aun concluded by mentioning questions that have been raised particularly around customization of alerts and messages. Customization—particularly, messages to drivers about specific hazards—is tricky, he explained, because, while messaging efforts should be led by federal agencies, other agencies and decision-makers have insights and want to have input as well. How best to alert drivers, and what to tell them in any given situation, are complex questions that highlight opportunities for further collaboration between the private and public sectors.
Philip Mai, Senior Researcher and Co-Director of the Social Media Lab, Toronto Metropolitan University, next spoke about his work with Anatoliy Gruzd, Professor and Co-Director of Research of the Social Media Lab, Toronto Metropolitan University, on developing social media research tools, such as dashboards and other visualizations, to further social science research on online misinformation. “Communalytic,” a portmanteau of “community” and “analytics,” is one such tool: “a no-code computational social science research tool for studying online communities and discourse.” This suite of data collection modules collects publicly available data from social media websites such as Reddit, X, YouTube, and others. The tool allows for multiple approaches to analyzing the data, doing what Mai calls “social listening” to see how information flows between users.
___________________
5 More information about Safety Cloud is available at https://www.haasalert.com/.
One area of focus is the coordination of crisis communication via social media. Coordination of this kind has become increasingly complex for several reasons, Mai noted. These reasons include the proliferation of new social media platforms, a scattered audience, a dearth of local news reporters, and the public’s declining trust in traditional news sources, which as a result are much less powerful and present than they once were. Mai and Gruzd have focused on understanding how communicators get their message across in this difficult information environment, especially if the message is urgent or otherwise timely.
One challenge to crisis communication via social media is platforms’ use of algorithmic filtering—that is, filtering what content individual users see and thus shaping their access to information—which dampens the likelihood of any single message going viral. Companies can pay to boost their messaging within this system, which also shapes what individual users see. Furthermore, Mai noted, social media systems are designed to privilege emotionally charged messages, and if an alert does not fit that criterion, it will not get shared as widely. Other challenges include misinformation generated by artificial intelligence (AI) and perpetuated by deep fakes; the disappearance of platforms and the resulting loss of audience; low levels of trust in information disseminated via social media; and, finally, the vulnerability of the system to state actors who weaponize it for their own ends (e.g., exaggerate, downplay, or otherwise misreport on a hazard or risk in order to discredit authorities’ responses). Mai and Gruzd called for further consideration about how much weight should be given to information found online and also emphasized the importance of supporting traditional media, which serves as an important alternative source for information.
The brief discussion that followed focused on the question of whether alerting drivers to various hazards in real time would cause people to panic and thus cause more accidents. Gerber responded that different types of messages for different hazards tend to have different effects. What action the driver should be asked to take will vary based on the particular hazard. Customization is possible, and the host platforms—Waze, Apple Maps, etc.—will shape the message the driver actually sees. Therefore, he noted, standardizing as much of the messaging as possible from Safety Cloud, before it goes out into the multiple delivery systems, is important.
RISK COMMUNICATION INNOVATIONS AND NEW FRONTIERS IN TROPICAL CYCLONE COMMUNICATION: PRIVATE SECTOR
The last panel of the day featured three presentations about innovations on risk communication in the private sector. Mike Chesterfield, Vice President, Weather Presentation and Data Visualization, The Weather Channel, shared some of the work being done by the Weather Channel around risk communication in the production of what it calls “immersive mixed reality”—a new technique in hazard visualization that uses hyper-realistic video imagery to depict a forecasted scenario. “We [are] able to show what the future’s going to look like in a video product.” Showing video content was deemed important, Chesterfield noted, because studies
found that visuals, including video, graphics, and images, increased the perception of risk, and that video in particular also lowered perceived uncertainty. In a way, he noted, these products “predict the future” in their depictions of “believable simulations” and in conjunction with on-camera meteorologists’ commentary. The presence of the on-camera meteorologist is important, he explained, both for the expertise they bring and to cue the audience to the fact that the video they are seeing is not real footage, but a visualization. These videos are meant to provide important context to forecasts otherwise shown only on a map, Chesterfield explained. One such product, Surge-FX, debuted in 2018, shows the NHC’s reasonable worst-case scenarios during an event (Figure 6.1). The message accompanying this product, and other such products, is that it is not a forecast, but a visual description of what people should be preparing for. Anecdotal evidence shows that this product and approach could influence how likely viewers are to take action in the event of an evacuation order or similar mandate, said Chesterfield, as shown in a recent in-house survey.
Flood-FX is another recent Weather Channel product that similarly depicts reasonable worst-case scenarios, in this case, by translating two-dimensional video into three-dimensional imagery. This product takes real footage of an area under threat and applies a simulation over it. These two products are intended to help viewers attend more closely to the forecast: “we hear all the time from on-camera meteorologists . . . [that] there is nothing more frustrating than having the forecast correct but people not listening,” Chesterfield noted.
SOURCE: Presented by Mike Chesterfield on February 6, 2024, created by the Weather Channel.
One important element of Flood-FX, Chesterfield said, is its capacity for “hyperlocalization.” Localized visualizations are a central goal going forward, he noted, describing how perhaps in the future, a product would allow users to enter their address and call up visualizations of what the forecast or various worst-case scenarios might look like in their front or back yard. “There is really no better way, in my opinion, to get people to react than actually show them what their future may actually look like.”
Another innovation Chesterfield highlighted is “metahumans, or personal weather assistants.” These are digital humans who would deliver personalized warnings to users—speaking in a way the individual understands, giving hyper-local information, and using their name, for example. These products are, he said, somewhat controversial, but they do represent the capacity of this technology.
Micah Berman, Lead Project Manager, Android Platform Safety, Google, described Google’s work to develop an earthquake early warning system (EWS) called Android Earthquake Alerts System (the System) for the Android phone platform, which is installed on about 3 billion devices worldwide.6 The System is a detection and distribution system, supplemental to the national EWS and other government warning systems, that uses the accelerometer already in these devices to detect and model an earthquake’s magnitude and epicenter in real time. This detection function is paired with distribution of alerts over a low-latency delivery network that is point-to-point internet protocol (IP) based. Berman explained that, at present, a bimodal model delivers one of two alerts, depending on the situation of the individual user. The first alert is a “be aware” alert, while the second is a “take action” alert. The two strands of action alert were developed to comply with official guidance set by multiple countries: a specific “drop, cover, and hold” instruction and a more general “protect yourself” instruction. These alerts are available in any of the languages supported by Android, Berman explained. They also are time sensitive and can be updated as the situation evolves.
Clicking through on an alert, users will see cached information about next steps. This feature is especially important in situations where power and connectivity may be unavailable. Information detected by ShakeAlert can also be pushed to the top of Google searches, so that details about the earthquake picked up by Android devices can appear in a regular search more quickly than information provided by official sources.
Berman noted that user feedback has been “extremely positive” and that areas of growth remain, foregrounding two questions: How to make earthquake warnings more actionable? and How to expand this capability to cover other types of hazards? Other possibilities for expansion might involve integrating with other capabilities already built into the phone, connecting people with information even if they are not in the locale (e.g., to allow them to learn more about loved ones who might be at risk), or giving a map and route in the case of evacuation.
___________________
6 More information about the Android Earthquake Alerts System is available at https://crisisresponse.google/android-alerts/.
John Lawson, Executive Director, AWARN Alliance, shared the alliance’s work in using ATSC 3.0, the NEXGEN TV transmission standard, to improve the delivery of emergency alerts and messaging.7 The warning system built on ATSC 3.0 technology differs from earlier ones, Lawson noted, distinguishing it from the Emergency Alert System (EAS) and WEA. The infrastructure underpinning ATSC 3.0 is TV towers, which are more resilient than cell towers. A battery-powered wi-fi dongle allows older television models to receive this signal and transmit over Bluetooth and wi-fi to other devices, he explained.
Lawson noted four important benefits of this delivery mode: the television signal provides geotargeting, which means that alerts—embedded in the television signal—will be received by a device that “knows where it is;” this locational information can enable targeted messaging. ATSC 3.0 also allows for rich media messaging that goes beyond text message alerts. Third, this technology enables most devices to be “woken up” if they are off but not unplugged; Lawson noted that this feature is very controversial. The final benefit is that ATSC 3.0 is the first and only internet protocol (i.e., IP)-based broadcast system for television. meaning it uses the IP for all types of information transmission.
In an ideal scenario around an “imminent alert,” the alert authority would issue a message that would be picked up by the station and distributed via geotargeting to individuals within the warning polygon. Users could then interact with this information, whether to dismiss or learn more: “We want to give consumers optionality.” Lawson noted that “interoperability” among alerting authorities and broadcasters is critical, explaining that, ideally, the alerting authorities and the broadcaster would at least enter into a memorandum of understanding “in terms of what they’ll push out.” Certain messages issued by alerting authorities would, ideally, go out “without anybody in the station touching it” rather than leaving the decision about whether the alert gets broadcast up to the news official.
At present, ATSC 3.0 is deployed across South Korea, Lawson said. In the United States, broadcasters are currently reaching around 75 percent of households that have televisions with an ATSC 3.0 signal. Outreach is critical to raising awareness of the existence of the NexGen TV/ATSC 3.0 signal and its potential for emergency messaging. AWARN Alliance developed some prototypes for EMs to use, and then conducted roundtables in which they could discuss the technology’s use. Broadcasters involved in the roundtables expressed strong interest in this technology. Currently, a pilot project in Washington, D.C., is under way. However, national leadership is currently lacking, said Lawson, but is fully necessary for a “widescale deployment.” These public-private relationships are critical for such a large-scale endeavor. Private solutions are in development as well, Lawson said, describing a start-up in which he is involved that would sell receivers that would bring advanced alerts, as well as a “disaster channel,” a 24-hour broadcast that would be programmed by AI drawing on public domain content.
___________________
7 More information about ATSC 3.0 is available at https://www.atsc.org/nextgen-tv/.
The question-and-answer session opened with a query from Jen Henderson about Eosco’s idea that social science researchers might partner with companies to use their apps and other technologies to collect data via surveys or other tools. Lawson responded that social science research is central to their work; Berman echoed Lawson’s enthusiasm for partnering with academic researchers. He noted that partnerships between large organization and researchers are sometimes not easy, but worth figuring out how to achieve.
An audience member offered two comments about “manufactured visuals.” First, viewers might develop a skewed sense of their safety because they do not realize that what they are seeing is not real and, second, trauma-informed communication practices advise against messaging that makes impacts explicit and that repeatedly names or describes impacts—which the Weather Channel’s products seem to be doing. Chesterfield responded that the Weather Channel takes great care to explicitly remind viewers that they are watching a simulation, in part by combining the visuals with narration that emphasizes this point, and to discourage imitators on social media from rebroadcasting images or videos of people who do put themselves in danger.
The final question came from Alex Lamers, NWS, who asked about the types of datasets that the NWS could gather and share that inform development of future innovations. Lawson noted that the roundtables revealed that “interoperability” was critical. Berman wondered whether, beyond earthquake detecting and alerting, other needs exist that the technology could address. He added that challenges around risk communication are not necessarily due to the data available from government sources; rather, they arise from gaps in the understanding of what users need and want, and what will compel them to take action. His concern, he said, is “less about the trigger and more about the experience that we built around that, and how we can make sure that is as effective and as integrated as it can be.”
