This chapter presents three approaches for transportation and emergency planners to consider in preparation for short term evacuation of part of a metropolitan area in response to an unanticipated harmful event. Here evacuation is treated as a temporary protective action. The assumption behind these approaches is based on the observations from New York City and Washington D.C. on September 11, 2001 that if a large magnitude harmful event prompts the evacuation of a portion of a densely developed downtown area the evacuation streams may include a large number of people on foot. The type of event visualized is that which occurs without notice; that is, without a long enough warning period to permit public safety and emergency response personnel to stage their resources in anticipation of an evacuation and for the downtown population to make preparations.
People may begin evacuating while the public safety and transportation authorities are still activating additional assets to manage the evacuation. People who are already in a private vehicle, along with those who can access their vehicle quickly, will create a sudden surge of vehicles. The tens of thousands of people exiting from the large downtown buildings in a relatively short time period likewise will create a large mass of pedestrians. The uncertainty accompanying the event, along with the disruption of the normal use of a given area, can be expected to create a brief period of seemingly chaotic behavior which then gives way into more purposive actions that will result in evacuees moving away from the area defined as dangerous. Law enforcement and transportation assets can be used to maintain a corridor through the city streets for the use of emergency responders going toward the dangerous area, prevent more people from entering the dangerous area, and enhance the mobility of the evacuees.
As described in Chapter 2, pedestrians apparently have not been an appreciable part of past evacuations in the U.S., unless they simply have been ignored in descriptions of evacuations predominantly characterized by the use of private vehicles. Much of the research on emergency response has focused on public response to warnings and what goes on around the event scene or in the command posts and emergency operations centers. Research on transportation tends to focus on the management of vehicle traffic. Transportation emergency planners recently have begun asking themselves if a large number of pedestrian evacuees are liable to have an impact on vehicle traffic flow, possibly jeopardizing their own safety in the process. Pedestrians, for better or worse, have far greater flexibility for choosing their route and expediting their evacuation than do drivers. This capability provides pedestrian evacuees with the potential for mingling with the traffic or interfering with the activities of responders while they pursue their collective objective of moving away from the dangerous zone.
For the sake of simplicity the most basic no-notice evacuation situation is used as the context for the discussion of evacuation. This is the situation of (1) a damaging event for which the impact has already occurred rather than being a threat with uncertain location and timing, and (2) which is assumed to be limited to the time and place of the initial point or points of impact; that is, there is not a plume of some dangerous material moving away from the point of impact. Other variables include time of day and weather conditions. We recognize that myriad permutations of these variables are possible which will alter basic planning assumptions for preparedness and the decisions made during actual response operations. All three approaches assume there will be appreciable numbers of pedestrians among the population making a temporary evacuation away from an unanticipated and major impact in the densely developed downtown area of a large city. The reader should consider this as mainly an assumption rather than something inevitable, there having been too few such events in large U.S. cities to know if the street scenes from the 9/11 attacks in New York City and Washington, D.C., are typical for those cities, or if they are likely to occur in other large U.S. cities.
The approaches outlined here are based on information from the review of the literature, ideas from anecdotal accounts of evacuations, and the views of practitioners and subject matter experts carried out as a preliminary inquiry into the topic. This is coupled with the research literature on social behavior in emergency situations and observations about transportation agency practices already used for accommodating large numbers of commuters or attendees at large public events in cities. At this time there is little systematic empirical data about the extent to which pedestrians might hinder the actions of drivers during evacuations, which of their actions are the most troublesome to response operations, and the extent to which they are responsive to instructions from authorities during an evacuation.
The three general modes available to individuals whether they are commuting from home to work or going about a city for errands or social engagements are: driving a private vehicle; riding in a private vehicle or public transit; or walking. People who are car-less in a large city normally have several options for making trips within the city or back and forth between home and work. These typically include one or more modes of public mass transit, and also taxis, specialized paratransit services, job-related shuttle buses, chartered or rented transportation for special purposes, and so forth. Many urban dwellers and workers also walk inform place to place in the compact downtown areas of large cities. As one reason, walking can provide more direct routes to nearby destinations than does mass transit, and others including that walking can be faster when vehicle traffic is slowed by congestion or that walking is enjoyable. All of these modes are suitable to use in situations that might prompt a temporary evacuation of a portion of a city. In densely developed urban areas evacuation in one’s private vehicle may not be the most effective for making a rapid evacuation. Under certain circumstances, greater use of other modes may be more suitable for achieving a balance of rapidity and safety for the largest number of people.
The penchant to stay with the familiar, the habitual, is characteristic behavior in many types of situations. Research into disasters and evacuations indicates it persists in emergency situations. Thus, some hunches are possible about U.S. urban evacuation behavior. In evacuation situations a preference for what is familiar contributes to that typically substantial proportion of the population that decides to stay put, even when warned of imminent danger, like a hurricane or flood. Most people who evacuate prefer to do so in private vehicles, if they have access to one. The workers who took mass transit to work won’t have a car in the city if an evacuation is ordered. Since they are already familiar with using mass transit to go home they are likely to consider that mode first. Evacuees whose choice is a regularly scheduled bus are likely to find themselves caught in heavy traffic congestion along with the evacuees in private cars. Rail transit can move large numbers of people quickly because it is usually grade-separated from road traffic. Workers in the downtown area who live outside the city are likely to want to go home, and so will readily evacuate the downtown area by whatever means. If necessary, they will walk. People who live in the area being evacuated, if they leave will go to the homes of relatives or friends if possible, rather than to the designated emergency shelters.
It seems safe to surmise that a large proportion of the masses of pedestrians captured in media photos leaving New York City and Washington, D.C., on 9/11 were motivated by the desire to get home. Family members tend to seek out each other under threatening circumstances. The evacuees appeared to be walking much further than was necessary to reach a safe part of the city. The exodus from New York City’s Borough of Manhattan during the power blackout across the Northeast U.S. in 2003 probably brought even more people onto the streets on foot because much of the mass transit system was disabled due to the lack of electricity. Most were on their way home from their place of work. In both cases the unanticipated event meant that public safety and transportation agencies were not appropriately deployed in advance in order to mitigate the initial problems caused by the sudden surge of traffic out of the city. In some places the pedestrians made their way through the congested and gridlocked traffic in a manner that alarmed transportation managers who view it as important to keep pedestrians off roadways both because they might impede emergency vehicles or the evacuation traffic and at the same time jeopardize their own safety.
The observations from the evacuations during the 9/11 attacks and the 2003 blackout indicate that large numbers of people evacuated from the Lower Manhattan area on foot as well as in personal vehicles. The large numbers of pedestrians in the evacuee stream may be an artifact of the high use of mass transit for commuting—53% in Manhattan and 33% in Washington, D.C. During the evacuation of downtown Cleveland on 9/11, when there was a possibility that a hijacked plane was headed that way, there was no noticeable pedestrian evacuation. People went to their cars, joined the rush-hour like traffic, and made an orderly if slow departure from downtown Cleveland. Compared to Manhattan and Washington, D.C., normally a relatively small proportion of Cleveland commuter trips are made on public transportation (12%). There was no mention of a much larger volume of people using public transportation when people began to leave downtown Cleveland on 9/11. This could be attributed, at last partially, to the fact that there was no real urgency and no damage that made it impossible for people to retrieve their cars.
In New York City and Washington, DC on 9/11, not all people who started out the evacuation as pedestrians completed their journey on foot. From anecdotal accounts and observations by transit officials many evacuees who started out on foot took advantage of some mode of public transport somewhere along the way. For example, they may have chosen their pedestrian route so they would arrive at a subway station, bus stop, or a ferry dock. Or as they walked, they may have noticed that the train or bus was an option. Some may have been offered rides by people evacuating in their personal vehicle. Both the traffic and transit managers are likely to be of the opinion that the pedestrian evacuees who switched to mass transit along the way enhanced their own safety and may have reached their preferred destination more quickly.
This suggests two overarching concepts for managing pedestrian evacuees. One is to determine the ways, if any, that evacuations involving large numbers of evacuees on foot as well as in vehicles call for special measures compared to the more commonly observed evacuation in the family car. One special measure that planners are considering is that of keeping vehicles and pedestrians separated as much as possible during the evacuation. Another is to apply traffic and pedestrian management techniques that result in reducing the number of evacuees on foot. The rationale for this is that pedestrian evacuees can avoid some of the hazards they might experience if they are accommodated through the provision of dedicated mass transit that can move them away from the dangerous area and well into the part of the city that was not directly affected. There the evacuees can engage in their usual and familiar actions for getting a ride from a relative or friend to their home, if they evacuated from their place of work in the city. Or they can find alternate lodging if they evacuated from their residence in the city. Two approaches related to this concept are presented here.
Moving a portion of the pedestrian evacuees onto buses or trains, besides removing them from vying with vehicle traffic for access to constrictions in evacuation routes (e.g., bridges or tunnels), can result in moving more of these car-less evacuees faster. And if not faster, the evacuees can proceed at least with less exposure to conditions created, for example, by extreme weather or a nighttime evacuation. For evacuees with poor health or a low level of fitness, the availability of a ride can make the evacuation trip faster and mitigate potential health consequences and injury from walking an appreciable distance.
Given the typical traffic congestion that occurs with a metropolitan emergency evacuation, commuter rail has the potential to be the fastest evacuation mode. To be an effective option, the commuter rail system must remain operational, at least along crucial portions of it, and not be subject to further threat related to the cause of the emergency evacuation. Security precautions may lead transportation authorities and emergency managers to suspend subway operation immediately, although trains might be put back in service after further threat assessment. Given the potential for subway and other rail transportation to expedite evacuation of large numbers of people, the benefit of keeping mass transit running to quickly move large numbers of people away from the evacuation zone may outweigh the level of risk, actual or perceived, associated with a highly uncertain event.
Commuter buses may also serve a role in pedestrian evacuation, but this is more problematic than rail if buses must use the same routes that other vehicles use to leave the city. Buses that follow their normal routes as an emergency evacuation begins will be subject to the same major traffic congestion and curtailed mobility as the private vehicle traffic. Even if normal bus stops and routes are still in effect, there will be no way for potential riders to predict when—or if—buses will arrive. Many people will prefer to keep walking, while others will endure long waits. Situations that delay pedestrians leaving the dangerous area leave them exposed to potential harm for longer periods of time.
Three approaches for managing pedestrian evacuation are presented below. Each of these is based on some combination of ideas from behavioral and transportation literature and from the discussions with transportation and public safety practitioners as part of this exploratory study. The theoretical context for these approaches is a no-notice high impact event that affects some part of a densely developed central city, under the assumption of the best case scenario (see 5.1). The three approaches share a common objective: to ensure the safety and mobility of pedestrians while minimizing the likelihood that they may contribute to evacuation traffic congestion. All three approaches assume that many evacuees decide to walk some or all the way to their preferred destination. This might be due to a lack of access to a private vehicle or to their decision about that mode of transportation not being the most suitable for evacuating quickly.
1. Designate and manage separate evacuation corridors for outbound vehicles and for pedestrians.
2. Provide dedicated evacuation transit hubs at the outer perimeter of the evacuation zone.
3. Provide “bus bridges” from where large numbers of people are emerging from the buildings to designated points at the edge of the area being evacuated.
In practice, vehicles and pedestrian traffic are normally separated, with vehicles on the street and pedestrians on the sidewalk. Keeping to the sidewalk is customary and normative for pedestrians, reinforced by social pressure and law enforcement. It doesn’t work perfectly, but well enough to serve the purposes of both drivers and pedestrians most of the time. If the assumption is that most pedestrians will collectively maintain this familiar behavior when evacuating from an urban area, unless some unusual situation exists, some predictions are offered about what authorities might expect.
For purposes of a rapid evacuation of a portion of a city, the large surge of both vehicles and pedestrians may make the customary separation more difficult to maintain. Once the evacuation pattern is established by law enforcement with transportation agency assistance, most vehicles and pedestrians will be going the same direction. The vehicles are likely to keep to the street, but pedestrians on sidewalks where capacity is exceeded are less likely to adhere to the norm for staying on the sidewalk. Gridlocked or slow moving traffic enables pedestrians to brave the street, which may be a prelude to many pedestrians taking to the street in order to go faster than the sidewalk crowd, thereby inhibiting the potentially faster progress of the vehicles.
In a large city, with many streets going in all directions, options are available for the emergency evacuation plan to call for physically separating the bulk of pedestrian evacuees from the vehicle stream. This can be done by establishing different, but basically parallel routes out of the dangerous zone. By ensuring that pedestrians have their own corridor, individuals or groups can evacuate at a pace that is comfortable for them and avoid interfering with the progress of vehicles or of each other. This approach reduces substantially the potential for pedestrians to be injured by vehicles or interfere with the progress of drivers. It also reduces the potential for personal conflict between drivers and pedestrians.
Several factors could adversely affect the successful implementation of this approach. It may be easier to hold vehicles to specific travel routes than pedestrians. Rather than the separate route designated for pedestrians, some of the pedestrian evacuees may attempt to use the sidewalks along the evacuation route for vehicles because of its greater familiarity, for example. However, as long as the pedestrians stay on the sidewalk, separation of streams is maintained although the flexibility for the pedestrians walking at different speeds or as a group is reduced. Furthermore, if vehicle traffic is moving at speeds of more than a few miles per hour, pedestrians are deterred from stepping into the traffic lane.
From the discussions and photographs of the evacuations in Washington, D.C., and New York City, there are clear examples of vehicles and pedestrians mingling in the same roadway space. Because it is unusual, it made news. However, not enough specific information was found to know if this occurred in very many places. Competition for the roadway was particularly the case when the evacuation route included a bridge, tunnel, or other type of chokepoint for traffic. Whether pedestrian and vehicle streams use separate corridors or share the same one, they will compete for use of a bridge that is part of the major, or only, route away from an area unless there are enough alternative routes to also designate at least one for vehicles and one for pedestrians. The one designated for outbound pedestrians may need to have lanes designated for inbound emergency response vehicles.
In Manhattan, for example, street options for separate vehicle and pedestrian streams are available in all directions and there are several bridges and tunnels for traffic leaving the Island. Working from the experience with the 9/11 evacuation, the NYC emergency plans now include the contingency of different bridges and tunnels designated for pedestrian, personal vehicle, and emergency vehicle use when the objective is to clear an area of the city quickly, without regard for the direction people would prefer to go. This means the decisions have been made in advance and can be immediately implemented as soon as law enforcement personnel can get in position. Some alteration of the designations may need to be made, for example, due to contingencies related to the impact zone, but the decision procedure will be quicker than if the law enforcement and transportation agencies have discussed the details of implementing such a plan in advance. For cities without a basic grid pattern in and around the downtown the designation of separate but parallel corridors may be more difficult. Most large cities will have some barrier such as a river or Interstate highway that result in chokepoints for traffic needing to be bridged across these features.
The greatest challenge for establishing the separate corridors is probably at the point where people begin emerging in large numbers from the downtown buildings in the evacuation zone. The evacuation of tens of thousands of people from large buildings at approximately the same time overwhelms the capacity of the sidewalks. Until evacuees have meaningful instructions on what to do next, large numbers are likely to mill around trying to find out what is expected of them or what others are going to do, even if it impedes vehicles in the area. The sooner it is made evident that there is an “appropriate” action to take in order to clear the area, the sooner some of the pedestrians will begin to move. Basic social forces will serve to prompt the bulk of the pedestrian mass to do the same thing. This is probably the best time for establishing separation of pedestrians and vehicles onto different streets. A major objective is to eliminate as much as possible cross traffic of vehicles or pedestrians while the bulk of the evacuees move away from the dangerous zone.
The establishment and use of totally separate corridors for vehicle and pedestrian streams may be the ideal of metropolitan emergency evacuation planners and public safety authorities. However, there are other adequate if not perfect solutions for keeping vehicles and people separate enough to maintain safety and mobility for both evacuation modes. For example, outbound lanes of a roadway can be designated for outbound vehicles, and the inbound lanes can be shared by incoming emergency vehicles with pedestrians using the sidewalk along that lane, facing the approaching emergency vehicles. Keeping to the right side of a center line is normative behavior for drivers and even more so if there is law enforcement presence. Behavioral research indicates that in the early stages of emergency situations most people tend to see themselves as part of the collective emergency activity, and a high degree of cooperative activity to reinforce the official enforcement can be expected. News media photographs from New York City on 9/11 show a major roadway with pedestrians clustered in a column along the left side and a few scattered emergency vehicles moving in the opposite direction. Photographs also show that at least one bridge also was shared in this manner, with pedestrians having all outbound lanes and only emergency vehicles using the inbound lanes.
For this approach, buses would be staged to take large numbers of people to designated areas. Dedicated buses are made available, but the staging area for boarding will be outside the zone being evacuated. The evacuees must walk to the buses, the distance being dependent on the size of an evacuation zone. The evacuees are then transported further out from the evacuated zone to one of several pre-designated locations, such as shopping malls, that have the services evacuees will need for personal comfort and making family contacts. From these locations individuals can arrange transportation to their preferred destination. The emphasis of this approach is to get as many people away from the dangerous area as quickly as possible by having them leave from wherever they are at the time of impact using the route most directly outbound from that part of the city
As part of this approach, there is the assumption that most people who work in the city will have driven into the city and will attempt to use their private vehicle for evacuation. Others who are car-less and able or who decide walking is going to be faster than driving, can walk the same outbound route as that designated for the outbound vehicles, although possibly on a different street that the outbound vehicles. Those on foot will have to go only as far as the staging area for the buses. Given the typically sprawling footprint for the residential areas of most U.S. cities, people who leave the downtown area on foot will need to arrange private transportation to get home or to some other preferred destination. The areas designed as the destinations for the bused evacuees should be selected to be easily accessible for the drivers of private vehicles coming to pick up an evacuee. From there the evacuee may have to drive back to a location at the other side of the central city, but this will be taking place in a part of the city where vehicles will be going all directions under normal traffic control.
The approach of using a dedicated bus service for ferrying pedestrians between two fixed locations is consistent with existing downtown evacuation plans that were described by representatives of two medium-sized cities, Cleveland, Ohio, and Charlotte-Mecklenburg, North Carolina, for this study. Each of these plans sets forth a basic concept for downtown evacuation by starting with a planning assumption that the impact and hazard area is basically centered in the downtown area, and evacuees can travel outward in one of several directions to get out of the hazard area.35 A map is drawn up to designate quadrants and several streets selected in each quadrant as best for outbound evacuation routes for vehicles. The Charlotte map also indicates pedestrian routes, some of which are the same as the outbound vehicle routes and some are on different streets.
The selection of the streets that are actually used during an evacuation will depend on the nature and location of the extreme event that precipitates the emergency evacuation. Traffic control personnel will direct traffic to the routes that are determined to be most usable and most suitable for effective evacuation and will control the traffic on these routes to prevent cross traffic that would slow down the outward progress. Each plan was developed through a multi-agency planning process that also includes the downtown business associations. Businesses are expected to participate in the periodic education of the downtown workers about the downtown evacuation plan and potential routes. To the extent that evacuation by private vehicle is possible, the planners assume vehicles will be the evacuees’ preferred mode. In an actual event the zone to be cleared is not likely to be exactly the same as that on the map, but the basic concept of driving or walking directly away from the dangerous area is still operable. It will be necessary for law enforcement officers or other traffic controllers to direct evacuees away from the impact zone to the pre-designated routes that are appropriate under the circumstances.
In short, the basic concept is that evacuees will either drive out or walk out of the area, using evacuation routes that are designed to minimize cross traffic. The plans are designed to accommodate both a typical vehicle evacuation and a mainly pedestrian evacuation. For pedestrians, each quadrant has a designated “transit hub area” located approximately one to two miles from the center of the downtown, where existing facilities in these areas can accommodate large numbers of people and a bus loading area. In an emergency evacuation, buses will be deployed immediately to these transit hubs. This approach is more efficient and removes potentially disruptive traffic from the site directly affected by the emergency situation. Pedestrian evacuees who walk to these transit hubs will be bused away from the downtown area to other designated evacuation facilities such as shopping malls, where evacuees will have access to such things as restrooms, telephones and food. At this point, it is up to them to find their way to their preferred destination, while avoiding the evacuated downtown area.
Compared to commuter rail, buses have a greater routing flexibility, including being routed around damaged or dangerous areas and to a larger selection of destinations. One frequent contingent use of buses is to provide “bus bridges” around portions of a commuter train route that is temporarily inoperable. Special purpose bus routes and service also are used for pre-planned special events (e.g., concerts, sports events) in order to reduce traffic congestion at the event facility. The bus service is provided between the event venue and a designated location beyond the more congested zone, as a “bus bridge” for people who don’t want to drive in the post-event congestion.
This concept of taking the buses to where the people are and the experience with the logistics of doing this held by special events planners, provide a basis for designing an evacuation shuttle service. To make this a safe option for the evacuees not using private vehicles, the emergency response command will have to have confidence that it is safe to have large numbers of people gathered near the impact zone and that there is an area suitable for waiting and then boarding buses as they become available. If that is the case, this approach might be particularly important to consider under circumstances such as bad weather, nighttime, or as the option for people who know they cannot walk very far due to health conditions.
The main objective of this arrangement in an emergency evacuation would be to gather up evacuees as they evacuate buildings and move them quickly beyond the perimeter of the hazard zone or beyond the most congested area. This is best accomplished if an area suitable for use as a staging area for the buses and pedestrians can be quickly determined and loading queues established. As people evacuate buildings, they can be directed to the staging area. This helps to alleviate the situation where evacuees mill around outside their buildings, waiting for some instruction about what they should do next.
Pedestrians can be guided toward the staging area where buses continually arrive, load, and leave, taking the riders along streets designated for the buses to a specified area where it is safe for them to disembark. At this point, the evacuees would again become pedestrians, but would then be in an area where they can collect their thoughts and proceed at their own discretion with respect to direction and mode. People who prefer to keep walking can do so.
Major logistical challenges must be addressed to implement this concept for ferrying large numbers of evacuees even a short distance. The primary drawback is that without advance notice, it will take time to get buses and drivers in place. One solution might be to redirect buses from the scheduled routes in the area and make more efficient use of them for the evacuation than keeping them on their normal routes. Personnel to carry out information, assistance, and control activities also will have to be assigned to the loading area. Another potential drawback is that service capacity might be quickly overwhelmed and result in large numbers of pedestrians accumulating at the assembly areas.
Evacuees will assess the situation from their own personal perspective and are likely to sort themselves into those who prefer to keep walking and those who need or prefer to ride to a safer place. Given the logistical demands for implementing this approach quickly enough to serve emergency evacuation needs, it must be designed in concept as part of emergency preparedness activities. The concept of operations will have to be worked out in advance, and in coordination with other agencies that must be familiar with it and prepared to support it. The decision to implement it as part of the overall evacuation strategy can be made as soon as the incident command determines the area is safe enough for this activity.
There are at least three other theoretically logical approaches; minimize evacuation by private vehicle; minimize evacuation on foot; and, minimize evacuation.
Many examples exist of traffic delays occurring when extraordinarily large numbers of people leave in private vehicles from the same general area at about the same time. The subject matter experts at the working group meeting for the study addressed the issue of how to identify and implement the most effective protective action for downtown populations during an emergency evacuation. If law enforcement and transportation authorities do not have sufficient advance notice of the need to deploy traffic control resources for a sudden emergency response, major traffic congestion is sure to ensue. In the early stages, at least, pedestrians are likely to make faster progress than vehicles. If evacuees are prohibited from using private vehicles for evacuation, the outbound pedestrians will have even more flexibility in the use of the roadway, with the exception of corridors for emergency response vehicles and other authorities. Under certain circumstance simply telling people to ‘run for their lives,’ so to speak, might result in the most lives saved. This extreme form of pedestrian evacuation, however, is not appropriate for several segments of special needs populations. Also, the ability to ensure the presence and coordination of enough law enforcement and other resources to enforce a prohibition of private vehicle use would present a major challenge.
The theoretically logical option opposite of minimizing vehicle evacuation is minimizing the number of pedestrians competing for the same roadway as vehicles. That is, the restriction could be on walking unless the destination was to a nearby transit option, so that all evacuation is carried out with vehicles. To some extent, this is the ad hoc situation for many evacuations carried out in small and medium sized towns. In a large city, with many car-less people, this might be difficult to implement. However, examples of some of the processes for accommodating large numbers of people with private vehicles and mass transit have been demonstrated in New York City and Washington, D.C. For example, in 9/11 around the bridge and tunnel chokepoints for leaving Manhattan, police requested that motorists provide rides to evacuees approaching these points on foot. A similar use was made of the excess capacity in private vehicles during the New York City December 2005 transit strike, for both in-bound and outbound commuters. In Washington, D.C., normal commuting patterns include commuters who routinely request rides with strangers.
The third theoretically logical concept for eliminating the problems and delays associated with an evacuation is to order that people stay where they are until some dangerous period is over. This “shelter-in-place” approach as an option already is included in many emergency plans, especially around fixed-site hazardous facilities. A circumstance under which emergency managers may want to instruct people to shelter in place is the confirmed or probable presence of a toxic or radiological airborne plume. Skilled responders will have a fairly good idea of where the plume is, what direction it is moving, and how long it will take for it to pass over a specific area; the layperson without the expertise and equipment to determine the path of the plume is better off to stay put indoors in a recommended location. For many types of releases, sheltering in place in substantial buildings can at least reduce exposure while a plume passes. If the plume path is contaminated, the authorities will have to determine when and how people can leave the buildings that were in the path. People also may be advised to forgo evacuation and stay sheltered if there is a credible threat of something dangerous about to happen, but no credible information on the location it will affect. The little research found on how well people comply with advisories or orders to shelter in place indicates that many say they will do so, especially if they have received education about sheltering as a protective action. The actual level of compliance will depend on several factors. One factor is whether people perceive the instructions to be clear with respect to what they are to do and for how long. Another will be what other people in the same vicinity are doing, sheltering or evacuating.
35 For the Cleveland downtown evacuation plan, see: <http://www.city.cleveland.oh.us/emergency/downtownemergencyevacuationplan/dowtownemerevacind.html>
For Charlotte-Mecklenburg plan see: <http://www.charmeck.org/Departments/Police/Home.htm> , Center-City Emergency Evacuation Plan.