___________________

¹ More information about OBXAlerts, a platform for alerts and notifications from Dare County and surrounding towns, is available at https://www.darenc.gov/departments/emergency-management/emergency-alerts.

² More information about IPAWS, a national system for local alerts, is available at https://www.fema.gov/emergency-managers/practitioners/integrated-public-alert-warning-system.

risk is always a challenge, but that continuing to improve communications skills is important and always ongoing.

Russell Strickland, Secretary for Emergency Management in Maryland, then discussed decision-making at the state level. People are the focus of all efforts, which are aimed at improving chances of survival, he noted, echoing Pearson. The state of Maryland works within a tiered system, in which planning, response, and recovery efforts unfold at the local, regional, state, and federal levels. “All events are local,” and thus, in this “across-the-board team effort,” the state’s primary role is to support local jurisdictions as they respond to threats and hazards in their communities.

Strickland highlighted three key aspects of the state’s role in decision-making around risk communications and emergency response. First, issuing information about and warnings for all hazards to the public is “a core capability in all phases of emergency management” (e.g., mitigation, preparedness, response, recovery). In Maryland, Strickland notes, this capability exists at both the local and state levels of government, the latter of which supports the former in communicating to hazards and threats. Second, the state coordinates communications and responses when individual jurisdictions are overwhelmed or when multiple jurisdictions are involved and affected. Finally, the state is responsible for delivering “coordinated, prompt, reliable, and actionable information [about a hazard or threat] to the whole community through the use of a clear, consistent, accessible, and culturally and linguistically appropriate method.” The goal is always to support local jurisdictions in their responses by “provid[ing] the right information at the right time to the right people so they can make the best decisions,” Strickland emphasized.

EM decision-making is shaped not only at various levels of government, but also by specific threats and hazards. In the latter context, Daphne LaDue, Senior Research Scientist, University of Oklahoma, and Tom Cova, Professor of Geography, University of Utah, discussed tornadoes and fire, respectively. LaDue spoke about her research into risk communication around tornados between city and county EMs, other local officials (e.g., fire captains, school officials, public works supervisors), and survivors of tornadoes. She stressed that most communities and individuals within a tornado warning area will not directly experience the tornado. Thus, individuals will often not take protective action until they actually see, hear, or feel a tornado coming; and EMs on the other hand, by the nature of their responsibilities, take action but also look to minimize the cost of being prepared. “The hard truth [for EMs] is that their jurisdiction probably won’t get hit.” Finances play into response decision-making, she said. Launching a tornado response is costly, and LaDue pointed to one community that had spent $200,000 in overtime and other forms of preparation but did not have a single tornado touch down. Past experiences also play a role in decision-making, especially as EMs seek to narrow down when and where impacts might happen within the 2- to 8-hour forecast window provided by the NWS. The timing of forecasts and warnings from the NWS also influence how much time EMs have to make decisions in order to get personnel in place if needed.

Research by the University of Oklahoma’s Institute for Public Policy Research and Analysis builds on LaDue’s studies, she noted, and includes a survey asking EMs to rank what type of information was most important at key points in a severe

ering critical and accurate data and passing it along to other local officials. The relationship between EMs and the NWS is strong, which is critical to the flow of accurate information and weather-related decision-making.

LaDue concluded her presentation with a brief discussion of gaps and opportunities in this area. She noted that demographic information of EMs as a group is lacking, as are data about job stability and turnover. In addition, the current forecast scale—which covers a relatively large geographical area—does not serve the small geographical area of concern for an individual or a city or county EM. It would be helpful to learn what forecast information means for them and use this knowledge as a basis for some level of autonomy as they make decisions for their communities. Figuring out how to support them in this learning and decision-making is important. Finally, there is a need to build greater resilience to severe weather, largely because many decisions are made in the final moments before a tornado strikes, and actions are not taken until the threat is clear.

Tom Cova focused his discussion on wildfire risk communication and the timing of decision-making and evacuation processes. Elements of timing include lead time (i.e., the duration between the trigger point and when the situation becomes dire); the length of time it takes for officials and then the public to make decisions; and the design of the community itself, specifically the paths of egress. Delays or mismatches in these elements of timing can make a situation dire; for example, in the Camp Fire of 2018, the time to evacuate exceeded the lead time, so the situation was dire from the start.

Exurban development at the periphery of metropolitan areas can cause longer evacuation times, Cova said, which can make a situation dire before people even realize it (Cova et al., 2021). Fires have also become more extreme because of climate change, which brings extreme winds and greater drought that can affect spread rate and flame lengths in unprecedented ways. Uncertainty exacerbates the problem of timing because it makes it extremely difficult to accurately predict the time available to evacuate. Cova emphasized how dynamic these situations are: the fire can spread more quickly than anticipated, or in a different way than anticipated and roads can close or become blocked. Cova also pointed to situations when little to no warning was given because warning systems were not activated. He refers to this situation as a “human problem” wherein a clear trigger point was not set, or it was not set at the right point. In these cases, challenges around the timing and sending of alerts has led to longer evacuation times in dire situations.

Echoing other panelists, Cova emphasized the uncertainty that arises particularly around unprecedented events. Models are often too optimistic, he said, and do not simulate dire enough situations. All of these factors—uncertainty, more extreme events, exurban development, and the need to simulate more dire events—highlight the “need to assess and improve protective-action risk communication.”

Social inequities also play a role in the effectiveness of risk communications. Jim Elliott, Professor of Sociology and Co-Director of the Center for Coastal Futures & Adaptive Resilience, Rice University, spoke about how social factors influence how people understand and respond to risk more than researchers might think. For example, a failure to evacuate is often explained as a transportation gap—that is, when

people do not have access to a car or cannot drive because they are too young or otherwise unable to drive. This transportation gap can be directly addressed by increasing access to transportation in areas with high concentrations of people without cars and by creating shelter-in-place plans that, perhaps, use public infrastructure for worst-case scenarios when people cannot evacuate. However, lack of a vehicle or access to transportation is just one reason why people may not heed evacuation orders, he explained.

Elliott’s research has revealed that a person’s decision to remain in place may be explained by three sociological reasons: religion, race, and roles (see Haney et al., 2010). For individuals whose religious beliefs explain extreme weather events as acts of God, remaining in place can be an act of trust in God. In this case, effective risk communication may entail working more closely with religious leaders, who may be more trusted sources of information than government sources. Such leaders could encourage their congregations to interpret and respond to such events in ways that align with and amplify government messages and mandates.

Racial inequities can increase exposure and vulnerability to hazards and threats in a multitude of ways, Elliott said. One less-discussed area of vulnerability is that homeowners of color often have lower wealth and less comprehensive insurance. In this case, an inequitable recovery process makes a quick and complete rebuild difficult, which in turn informs planning before and during a storm. Owning a home creates a strong incentive to stay and repair damage as quickly as possible. Concerns that a small problem will become a big problem if not fixed immediately can lead to a person remaining in place rather than facing of potential for leaving their home and being unable to return. “People sometimes stay precisely because the risk that’s communicated creates a lot of worry about what’s going to happen to their home and their economic future,” Elliott noted. Situations such as these are particularly common in historically marginalized communities, where not only are resources scarce but also trust that government assistance during the recovery will be adequate is low. Elliott suggested that government communicators could help to address this issue by engaging more with economically vulnerable communities, particularly in areas with high homeowner rates, where people are incentivized to stay and protect their financial investment.

Household roles also influence how people respond to an evacuation order. Elliott noted that often, traditional gender roles come into play, with women and children evacuating and men staying behind to protect the property. In addition, household earners who are worried about losing their jobs often stay behind. In these cases, engaging with employers, particularly those who employ “hourly workers in traditionally male sectors,” could help to improve response to evacuation orders and similar types of messages.

Messaging about the hazards themselves is not enough, Elliott noted. The cascading and compounding effects that impact the social and built environment also must be communicated to the public with an eye toward social factors. He cited the example of road closures due to street flooding, which can cut people off from important services even though they are not in a flood zone. Another example is the threat posed by hazardous industrial pollutants that could be released if infra-

structure is damaged during a storm. Elliott suggested that people responsible for risk communication might prioritize areas threatened by both chemical and natural hazards—for example, by engaging in early and ongoing communication with local leaders to raise community awareness about how the mobilization of hazardous chemicals via local storm waters can make remaining in place more dangerous than it might seem based on the storm forecast alone.

Fugate, as moderator, opened the question-and-answer portion of the session by observing that there is a legal distinction between warnings, which are issued by the NWS, and evacuation orders, which can only be issued by state and local jurisdictions. Therefore, he wondered, who makes these decisions, and what are their sources of information around those decisions? Pearson explained that in North Carolina, the authority to issue an evacuation order, declare a state of emergency, and impose various prohibitions or restrictions lies with the governor, who delegates to county officials, who then may delegate to municipal officials. Thus, one challenge for local EMs is drawing together the NWS forecast with local knowledge and ensuring that elected officials at the county and local levels have correct, critical information as they decide on the course of action. State-level decision-making resembles the local level in terms of briefing officials in order to help them make informed decisions, Strickland noted. Discerning what information officials are looking for as they make decisions is critically important, he noted. In Maryland, he explained, in the event of a storm, the state names an evacuation coordinator who works closely with officials (e.g., county manager, mayor) in the local jurisdictions that are most likely to be impacted, and across jurisdictions as well, and keeps the governor apprised.

Rumors and misinformation—whether deliberate or unintentional—pose significant challenges to risk communication and response, and were the topic of Fugate’s second question: How can rumors be addressed and controlled? Listening to what sort of messaging and information is circulating is key to fighting misinformation and rumor, Pearson said. His county establishes a call center with a hotline that people can call to receive accurate information and report any rumors that are circulating. With this knowledge, county officials can directly counter misinformation with accurate facts from a trusted source. However, Pearson stressed, officials must be tuned in and listening to what the public is communicating. Strickland added that similarly, in Maryland, smaller jurisdictions establish rumor control hotlines, which feed into the Joint Information Center, which then verifies information to send back to the jurisdictions.³ Elliott noted that a large part of rumor control for his jurisdiction is coordinating consistent messaging among the many agencies commonly involved. In addition, alerting local news outlets to prioritize or use information coming from a central source, rather than other sources of potential misinformation, ensures that they are “a partner in these conversations.” Because many people turn to the local news when storms and other hazards threaten, coordinating with local news outlets can be especially helpful, he said.

Castle Williamsberg asked the final question, “What is the single-most important operational challenge faced by practitioners in effectively communicating

___________________

³ Information about a Joint Information Center from Washington County, Maryland, is available at https://www.washco-md.net/jic/.

weather-related risks? It is a challenge to keep the message simple, Pearson said, but it is essential to be “clear, on point, and unambiguous.” Strickland agreed, adding that balancing uncertainty and confidence in the midst of a changing forecast is also challenging. Giving people accurate, meaningful information with enough time for them to respond can be difficult. “Understanding what people are responding to” can be essential, LaDue added, which might include awareness that multiple warnings or hazards (e.g., hail and winds) are driving behavior, depending on which is prioritized. She stressed the importance of understanding how the communications and decision-making systems work: a warning from a local official (e.g., a sheriff) may feel more directed and relevant than one from the state to a larger or more general area.

ROUNDTABLE: COMMUNITY LEADERS AND COMMUNITY ACTION

The session on risk communication and decision-making in communities continued and concluded with a discussion between Jeff Lindner and Archie Chaisson, Parish President, Lafourche Parish Government, Louisiana, chaired by Brad Colman (committee member), President, American Meteorological Society. The two discussants began by talking about the information they consider when making decisions about risk communication in the event of a storm, and what information they would like. Lindner stressed the importance of “primary sources”—the NWS and the National Hurricane Center (NHC)—in gathering information with which to brief officials during a hurricane or other event. The information from these two sources was important, Lindner said, in guiding decision-making during Hurricane Laura (2020). While TV meteorologists and European weather models made more disastrous and different predictions, the NHC’s forecast “really never changed for about 48 hours,” and the NWS expressed a high level of confidence in that forecast track. In this case, basing decisions solely on information from the NWS and NHC resulted in an appropriate level of response and not an overreaction. “It’s just as important sometimes not to pull the trigger as it is to pull the trigger” Lindner noted, especially when dealing with a population size in the millions, which can be the case for storm surge impacts and warnings.

Lindner highlighted the mismatch between various milestones on the timeline for hurricanes at which decisions must be made (to allow enough lead time) and the timing of when crucial information is available. Ideally, he said, the lead time on the forecasts from the NWS would be longer. Because of the current mismatch, decisions must be made even when the level of uncertainty in high: for example, in Harris County, Texas, mobilizing busses for use in an evacuation takes about 96 hours, and thus the decision to order this resource from the state must be made 96 hours out, when uncertainty is often high, if it is to be effective.

At 60 hours out, Lindner explained, officials on a multi-agency call decided to evacuate the general public, a process that involved coordination among counties or parishes, including, for example, a phased approach so that coastal counties issue evacuation orders several hours before more populous metropolitan areas such as

Harris County, so that the coastal population can clear out before the traffic increases. This approach is especially important, he noted, in high-population areas such as the Texas coast.

In addition to population density, lag time between order and response is another important factor to consider when contemplating the timing of messaging, Lindner added. There can be a 12- to 24-hour gap once decisions about evacuation become public. Closing the schools once the evacuation order has been given can help shorten this lag time, he said.

Uncertainty can profoundly impact the decision-making process, including questions of how to communicate uncertainty itself to stakeholders, whether that be the public, elected officials, or the media. Chaisson commented that uncertainty often looks like preparing for something that does not actually happen—i.e., “crying wolf.” This situation can happen repeatedly, as in the storm season of 2020, and thus skew people’s sense of risk in the event of a truly threatening storm. Similarly, people will remember a benchmark storm that they survived and disregard the present threat. These two groups of people are more likely to remain despite evacuation orders, and it can be very difficult to effectively communicate uncertainty to them. Chaisson referred to Hurricane Ida (2021), which was an unprecedented storm for the area, and described how he and others worked to communicate that this event would in fact directly hit the area. He noted that because Hurricane Ida was more of a wind event than a water event, the impacts were not as catastrophic as they could have been. Lindner briefly added that a confident forecast from the NWS or NHC helps to remove uncertainty, which is reflected in more strident language used by both the NWS and the local EMs.

Costs associated with evacuation orders often impact decisions about whether or not to issue them, especially when uncertainty is high, Colman noted. Cost is secondary to saving lives, Lindner stated. However, the dampening effects of alert fatigue are problematic as well, and striking a balance between the importance of saving lives and the necessity of avoiding alert fatigue can be difficult. Normalizing situations where “nothing happens” can be helpful, he suggested. Chaisson noted the way that cost factors in the decisions of individuals who cannot afford to evacuate: “they don’t have the means, and they won’t leave for nothing.” Certainty has to be high for this population to evacuate.

Colman stated that “environmental equity,” a term used in relation to populations who are more vulnerable because of their location and/or lower income, is also an important factor in decision-making in the moment, as well as for the future. Chaisson noted that, in their parish, evacuation protocols and other plans are based on proximity to the coast; the population deep in the parish, closest to the coast, is environmentally more exposed to threats and experiences a higher rate of poverty. Evacuation often involves moving people up the coast and parish little by little, rather than all at one: people will first go up to family and friends, and then slowly “migrate up,” often staying in hotels or schools serving as temporary shelters. This approach differs from mass evacuation, which was necessary for Hurricanes Ida, Ian, and Harvey, which brought flooding to densely populated areas and necessitated

chartering busses and erecting mega-shelters.

Educating the public is key, Lindner added. Education can involve working closely with local leaders, especially church leaders, before storms to help people understand how to navigate the evacuation system. It requires clarity in the messaging about the type of evacuation response necessary given the various threats and hazards. Lindner cited the example of people believing they need to drive very long distances to escape a hurricane when, in fact, driving a much shorter distance to escape storm surge flooding may be all that is necessary. Distributing information ahead of time is good, but the underlying challenge is that “no one really pays attention to this until it’s happening.” Chaisson and Lindner both reiterated the challenges to communicating risk to people who have experienced evacuation orders when “nothing hit” and to people who stayed and survived bad storms in the past—and, in both cases, downplay the risks of the present threat.

Chaisson noted that decisions about evacuation are never simple or easy. He particularly stressed the emotional difficulty of ordering an evacuation—essentially asking a person to leave everything they have with no assurance that they will be able to return to it, or that it will be there if they do. It is “probably one of the most difficult things you will ever have to do.” He emphasized that often leaders need to trust their gut, prior experience, and the plans already in place.

HIGH-LEVEL SUMMARY OF SESSIONS THREE AND FOUR

Jeanette Sutton, Craig Fugate, and Brad Colman provided a high-level summary of sessions three and four. Themes and concepts that cut across the two sessions—and resonated with the broader workshop discussion—included the need to use plain language and avoid jargon; the crucial role of partnerships in deepening understanding of the various stakeholders’ needs; and the importance of considering those needs when building and implementing a communications strategy, crafting messages, and making decisions. They acknowledged that any given audience or population is not monolithic; therefore, communication strategies would ideally reflect audience diversity.

Sutton noted that discussants reinforced the concept that messaging works best when it also addresses a specific audience. Here, messengers must understand who the members are, what they need, and how they understand different types of information. Whether that means a map, a focus on impacts, or an emphasis on vulnerabilities, attention to the audience’s greatest needs complements the use of plain language. One example of an audience requiring tailored messaging is people who are undocumented, given current political realities. The two-sided nature of communication—message and receiver—was raised in discussions about new technologies and products, which could be created and assessed with audience needs in mind.

Although the hazards discussed by the breakout groups—earthquakes, extreme heat, and flooding—vary greatly in terms of speed of onset, lead time, interaction of compounding hazards, and partner organizations that can serve as intermediaries between government agencies and the public, they offer common themes in terms of

risk communication. One shared theme is the importance of using plain language, rather than jargon or overly technical language. Fugate shared the example of the term “storm surge,” which is often not accurately understood by the public. The “tendency for precision” can be a hindrance in helping individuals understand how to act.

Fugate, reporting on the first set of panels from session four, echoed Sutton’s highlighting of attending to the different needs of the community members that comprise an audience. In this, he captured a point made throughout the day: that messaging often works best when it addresses a specific audience. This involves understanding who that audience is, what they need, and how they understand different types of information. He highlighted one phrase— “We trust the forecasters, not the forecast”—as being particularly revealing of how the general public responds to uncertainty. He raised the question of how to communicate particularly with people who are undocumented, given current political realities.

Another audience, as it were, is EMs; Fugate highlighted the discussion around the need to better understand how and why EMs decide to activate or not activate warning systems, especially when the situation involves high levels of uncertainty. Coleman noted that EMs approach the decisions and pressures they face with a great deal of preparation, ingenuity, and resourcefulness, helping them to be focused and decisive in the moment. Discussions also highlighted the great effort and wide range of resources that can be used to mobilize an emergency response. In all of this, Colman noted, strong partnerships—between local EMs and the NWS—help EMs to filter out the noise. Finally, EMs sometimes advise, or are themselves, elected officials; therefore, their decisions could become an election issue—which resonated with the workshop-wide focus on the local dimensions of information and communication.

Partnership emerged as another prevalent theme and spans connections across the public and private sectors (especially around technology), as well as intermediaries such as community organizations that can help connect researchers and government agencies with locals and public health organizations.

This page intentionally left blank.