Under disaster conditions, providing information to the traveling public is urgently needed to maintain the safety of travelers in the disaster area or to divert travelers headed toward the impact area so that they can continue toward their destination, although by a less than direct route. When an area is threatened and evacuation is called for, information needs to be disseminated that prompts people to travel out of the area as soon as possible and along safe routes. Ineffective communications can increase the potential for harm to the public.

While public agencies involved in emergency response rightly view their mission as protection of the public, a common assumption is that people will take the appropriate action when the authorities provide information and recommendations about the emergency. Actual experience with dangerous conditions has shown that an appreciable portion of the target audience does not engage in the recommended action. Some part of the intended audience may not have received it, delayed their response to the recommendation, or chosen to make an alternative response to the recommended one. On the other hand, there may be people who were not part of the intended audience who make the recommended response even though it is not relevant to where they are in relation to the danger. The failure on the part of the public to heed disaster information can result in death and injury.

Behavioral Research Basis for Disaster Information Dissemination

The dangers to the public posed by ineffective communications during disasters has prompted researchers to examine the human response to actual or potential dangers so that methods of communication and the messages conveyed can be adjusted to achieve a higher level of the recommended response to disaster conditions. These research findings provide an important foundation for public sector strategies for communicating with the public during disasters. Pertinent findings include the following:

• People prefer to try to continue their routine and familiar activities as long as possible if they are at home. Many people will interpret an emergency warning as not being pertinent to them or the area where they are, or perceive that the situation being described does not necessitate a response such as evacuation or rerouting around the area to be avoided. Thus, people may stay where they are when evacuation is prudent, often citing their previous experience with a warning that indicated it was not necessary to evacuate or that they did not want to get caught in the slow moving evacuation traffic.
• Normal social behavior is to take actions to protect one's immediate family or other relatives. When information is disseminated that indicates an emergency condition or the need to evacuate, people will begin to move about the area in an effort to get their relatives together. They are unlikely to evacuate until this is accomplished. This can both delay the departure of many people and create some degree of traffic congestion. However well crafted a message to evacuate is, it is not likely to minimize this behavior. Emergency responders and emergency managers need to recognize that this will occur and make plans to accommodate it.

• Audiences are not homogeneous. Research on the use of information typically finds that different segments of the overall audience receive, process, and act on information differently. Variables that have been found to affect responses to information about hazards or warning messages include education level, economic status, ethnicity, and previous experience. This makes the task of providing meaningful and persuasive messages to all audiences challenging.
• Panic is not the normal public response. Research on emergency response has found evidence that many law enforcement personnel and elected officials, among others, hold the misperception than people confronted with information about a threatening or dangerous situation will panic and begin to engage in behavior that threatens their safety and that of others. News accounts may reinforce this perception by their choice of words and situations depicted out of context. The danger is that authorities who continue to hold to this belief may withhold critical information that could help people avoid injury or death. Transportation authorities need to understand that it will be easier to predict and facilitate people's withdrawal from a dangerous situation when they have been told the characteristics of the threat and what their risks and options are, than if all they know is that there is some danger.
• Considerable research on social behavior in evacuations and disasters substantiates the ability of people to use information in the face of threat. People are frightened in these situations, but being frightened does not preclude rational behavior and, in fact, may be an important motivator for prompting people to take appropriate protective action. In response to their own reading of the situation or official warnings, people will make decisions that are rational in terms of what they think they know about the situation, even though in retrospect their information may be proved to have been inaccurate.
• Information seeking is the norm. Studies of evacuation behavior find that the most typical action people take upon receiving a message that recommends evacuation is to seek to confirm the first information they hear. Many continue to check multiple sources of information for the purpose of making sense out the situation, so they can engage in protective action. They will use whatever information they perceive to be most credible. This is true for both people who evacuate and those who don't. Accurate, timely, consistent, and specific information from officials on where the danger is and what the options are for moving away from it, such as routes and staging or shelter destinations will be shared with others and acted on. Timely provision of consistent information about what the authorities think is the safest course of action will help to avoid confusion and maintain calm among evacuees. A related finding is that the more unfamiliar and more imminent the threat is perceived to be, the more likely people are to evacuate and to do so fairly quickly. Thus, the public may begin to evacuate on their own in advance of any warning or recommendation by public safety and transportation officials. The implication of this finding is that officials need to be prepared to facilitate this action once they realize it is occurring.
• The behavior of response organizations themselves has been studied to figure out how to make them more effective. The biggest error that emergency response personnel tend to make is to believe that disasters are just big emergencies and that their normal practices for incident response will be effective. However, disasters create communication demands that are an order of magnitude above that appropriate to routine emergencies. Agencies must acknowledge this and develop appropriate protocols, including how to handle communications to the public. Also, most leaders of emergency response agencies acknowledge that the greatest value of engaging in a multi-agency planning process is not the creation of the document called the emergency plan, but the development of relationships and trust among agency personnel that facilitates collaboration when disaster strikes.

Behavioral research such as the findings cited here has bearing on the use of ATIS, the subject of this document. ATIS, like other methods for disseminating disaster-related information, should be grounded in an understanding of how people receive and act on information about disaster travel conditions so that ATIS-disseminated information can be most effective.

Assumptions and Constraints Affecting Use of ATIS During Disasters

Underlying this document are a few assumptions that bear mentioning at the outset. The extent to which those assumptions are not valid represent potential constraints on the efficacy of ATIS assets during disasters.

An assumption throughout this document is that transportation agencies have or are planning to have various ATIS assets that can be used to convey messages in near real-time about the condition of the transportation system and whether or not it has been affected by the disaster agent. While ATIS is generally widespread, there are gaps in availability.

An assumption is also made that communications assets in general are available by which information can be transferred that ends up as an ATIS-based message. Both of these assumptions need to be examined by agencies in the context of a specific region or specific disaster scenario. Several agencies, because of their particular response role or the location of their operations with respect to the overall disaster area, will have information that could be of value to evacuees or other travelers. Each of these agencies can be viewed as part of a supply chain of information for travelers, and a gap in the supply chain may result in important information not being transmitted. For example, an individual wildfire fighter might be unable to maintain contact with the incident command, because his communication mode is dependent on a line of sight pathway or being within a certain radio range. A county law enforcement agency may not be able to directly communicate with a city law enforcement agency or with the emergency operations center in an adjacent county because of a lack of interoperability among agency communication systems. Information about rapidly changing conditions that would be critical to travelers in a particular area may be lost or delayed due to such gaps in pathways for the communication of information from the field to the transportation agency for dissemination via the ATIS.

Another assumption is that public agencies have pre-planned protocols for communicating with each other and sharing information during emergencies. The lack of such protocols negatively impact incident response under normal conditions. By definition, disasters involve many more agencies and jurisdictions than is the case for a routine emergency; and protocols for an expanded set of agencies to be included in notifications and status updates need to be designed in advance of a disaster, along with the decision criteria for when to invoke the disaster communication protocol. Some areas may have experienced major damage to the roadway that is part of a critical transportation route, but the response agencies may not think to notify agencies beyond the impact area. Even in situations of widespread damage, the lack of information from a specific jurisdiction is likely to be interpreted as the lack of any problem there, rather than the lack of communication coming from there. For example, during the Loma Prieta earthquake in California in 1989, damage to communication systems resulted in a delay of information about damage in and around Santa Cruz that in turn resulted in a delay in response to the damage and information about damage to the major highway into Santa Cruz. There was no protocol for systematically checking with each jurisdiction, since under normal circumstances if a jurisdiction had a problem, it would have been reported. Thus, if a traveler is accustomed to receiving current information on roadway conditions through ATIS and the message is not current, the traveler is likely to assume there is no problem with that route despite the disaster.