On Aug. 29, 2005, Hurricane Katrina made landfall over Louisiana and Mississippi as the costliest hurricane in U.S. history. In the days before the storm, it is estimated that more than 1 million people evacuated south Louisiana.
Based on a preliminary review of traffic volume data recorded during the evacuation, it is apparent that more people were able to be moved over a shorter period of time than was ever thought possible. One of the primary reasons for this apparent success was the evacuation traffic management plan developed and implemented by Louisiana officials only months before Katrina’s arrival.
The southeast Louisiana evacuation plan was unprecedented in terms of its scope and level of management. When fully implemented it included the reversal of nearly 100 miles of interstate freeways between Louisiana and Mississippi, the closure of nearly 100 more miles of interstate and the rerouting of traffic on several other freeways and primary arterials. This article presents some of the traffic impacts from Hurricane Katrina in Louisiana and discusses some of the lessons learned from this and other similar mass evacuations over the past decade. It is hoped that this information will help to show how large-scale traffic management plans can be applied to improve evacuations in other locations, for other threat scenarios or even routine conditions or event scenarios where short-term traffic surges can be improved.
Years in the making
The Louisiana evacuation plan that was used for Hurricane Katrina was actually developed over several years, based on the experiences of two major evacuations of the city, Hurricane Georges in 1998 and Hurricane Ivan in 2004. Although no evacuation is ever wanted, these two events were valuable to test and refine the strategies ultimately used for Hurricane Katrina. In the Hurricane Georges evacuation it was clear that a conventional utilization of the highway network around the region was not going to be adequate. As a result, a plan to implement two short segments of contraflow was developed by the Louisiana State Police (LSP). Six years later when Hurricane Ivan threatened another direct hit on the city, numerous deficiencies in the post-Georges plan were evident. Most critical were the persistent over-reliance on the westward movement of evacuees, an inefficient contraflow loading strategy in New Orleans, extreme congestion resulting from the confluence of multiple regional evacuation routes in several cities in the region and an inability to access real-time traffic flow data and then communicate timely and accurate travel information to evacuees.
After considerable public criticism following the Ivan evacuation, the Louisiana Department of Transportation and Development (LADOTD) and the LSP formed a Louisiana Evacuation Task Force that included consultants from industry and academia to identify where and how the congestion occurred and develop and test ways to reduce it. The plan that was developed was truly ambitious and considerably more active than most other plans of its kind. Using a fast-track development schedule and taking advantage of the close partnership with their counterparts in Mississippi, Louisiana officials were able to finalize the plan and communicate it to the public just prior to the start of the 2005 hurricane season.
The fundamental objective of the Louisiana plan was to move people out of the most threatened areas of the metropolitan New Orleans area as quickly and efficiently as possible. As has been well publicized, the ground elevation of much of New Orleans is below sea level, and the area is protected from flooding by an elaborate set of levees that ring the area to keep flood waters from inundating the city. At its core was a philosophy to maximize the limited number of high-capacity routes out of the city and dictate the movement of traffic on these routes. Rather than giving evacuees travel choices while en route, their movements would be dictated by the LSP. Evacuees moving west across the Gulf Coast were not permitted to access or traverse freeways within the critical New Orleans evacuation area. Through the use of contraflow, outbound capacity could be significantly increased on several freeways. By controlling access and movements on a regional level, the congestion-causing intersecting and merge locations were effectively eliminated.
The first hours
Soon after the storm, efforts were started to assess and evaluate the extent of traffic impacts associated with the evacuation, including how, when, where and how long it took evacuees to move the road networks of the Gulf Coast, how contraflow affected the efficiency of the evacuation and in particular how some of these impacts may have been related to the revisions in the plan. Although the complete results of these analyses are not fully realized, early indications are beginning to reveal several key facts, such as the number of vehicles involved, key times of when the evacuation started and ended at key locations, the geographic extent of the evacuation in Louisiana, the directions of travel, the capacity gains made through the use of contraflow and some of the implications of the modifications made to the Louisiana evacuation highway management plan.
Although it is obvious that any evacuation of the magnitude of Katrina will have far-reaching and long-lasting effects on traffic, the extent of the impacts from this event were nevertheless quite surprising. Traffic count data recorded as part of the LADOTD traffic-monitoring network showed that traffic volume increased at every one of the stations for some duration during the evacuation period. Obviously, the most drastic increases were in areas closest to New Orleans and the storm landfall. However, volume impacts were observed on even lightly traveled roads, through sparsely populated areas, hundreds of miles from the storm landfall location. Increases in volume were evident as late as 60 hours after the start of the evacuation as evacuees were finally moving through the distant location of Monroe and Shreveport.
During the peak 48-hour period of evacuation, more than 430,000 outbound vehicles were recoded on the six freeways and major arterial roadways in the southeast Louisiana region. This total does not include several other primary highways that were not part of the LADOTD traffic monitoring network. Temporal analyses show that the actual thrust of the evacuation in the New Orleans vicinity lasted for a period of between 34 and 38 hours. If it is assumed that about 80 to 90% of the population evacuated, the duration of the evacuation is merely half of the 72-hour clearance time that had historically been assumed for the area prior to the 2005 plan modifications. This was a tremendous accomplishment given the limited number of roadways and made even more significant by the apparent lack of traffic-related injuries or deaths directly as a result of the evacuation.
The preliminary analyses of the evacuation also reveal other indications of the effectiveness of the plan. Most noticeable was the impact of contraflow. By themselves, the two northbound contraflow freeway routes into Mississippi (I-55 and I-59) carried nearly 160,000 vehicles or more than 35% of the total recorded volume out of the New Orleans area. On I-55, outbound volume during Katrina was increased by nearly 40% to about 24,000 vehicles over a corresponding 48-hour period during Ivan when contraflow was not used on this route. At an occupancy rate of 2.5 persons per vehicle, this increase translates to about 60,000 more people. The data also illustrated the travel patterns on secondary highways. The data suggest that during Katrina many of these routes were well utilized by evacuees seeking to avoid more heavily congested primary routes. If ultimately true, this observation would support another of the major objectives of the evacuation plan: to maximize overall system capacity by dispersing traffic to all available routes rather than concentrating evacuation travel on the interstates.
Can’t save everyone
Not all the evacuation news in Louisiana was positive. Among the most indelible images from the Katrina tragedy in New Orleans were the scenes of people picked from rooftops by helicopter, unable to move from refuges such as the New Orleans Convention Center and Superdome and waiting on freeways for rescue. Despite the clear successes of the highway-based evacuation plan, these accomplishments will always be contrasted against and overshadowed by the inability to get everyone out of harm’s way. How and why not everyone could or chose to evacuate before the storm may never be answered completely.
In most every evacuation there is some percentage of population who, while having the capability to evacuate, do not leave. It has been estimated that between 100,000 and 300,000 people did not evacuate the city. The majority of these individuals were likely neither daredevils nor thrill seekers. Rather, their refusal to leave was likely founded in much more practical concerns such as the need to care for a loved one who may be too frail to be moved or the desire to maintain the soundness and security of their personal and commercial properties. For some, it has been suggested that a recent history of “close calls” requiring unneeded evacuations left them with a “Cry Wolf” attitude toward evacuation orders.
On the other side of the nonevacuees issue were the individuals with little or no means of travel. Despite their desire to evacuate, many individuals did not have the capability to move themselves. These groups, often referred to as “low-mobility” populations, included the infirm, the elderly and the economically disadvantaged as well as people like tourists and the incarcerated. Newspaper stories pointed to census data that showed that about 112,000 did not have access to a personal vehicle at the time of the storm. Since it also was well recognized that a substantial portion of New Orleans residents did not own their own vehicles or have ready access to transportation, plans for moving these various groups of individuals were in place and they were implemented prior to the storm’s landfall. How well they were communicated or utilized is an ongoing subject of debate.
One of the most obvious targets of criticism were local governments who received blame for not planning well enough or providing adequate transportation to shelters of last resort. Realizing this problem and the limited capability of the local government to move such an enormous number of people in short order, local public officials have long advocated “neighbor helping neighbor” policies by urging low-mobility individuals to seek evacuation transportation with friends, family, neighbors and church members.
Despite the preparations and efforts of local officials as well as the reluctance on the part of some to leave, there are those who thought more could have and should have been done. Most important was the issue of communication. Evacuation plans are only effective if people are aware that they exist, and evacuation orders can only be heeded if they are heard or followed if they are received in time. However, one of the clear messages to come out of Katrina is the need for personal planning and self-reliance in evacuations. The sole reliance on the government for one’s personal health, safety and welfare will always be a shaky proposition, despite the best planning and intentions.
Needs some work
One of the difficulties in assessing different evacuations is that there is so much variability between events, including the mindset of the evacuees, the perceived level of threat and the location. However, in reading and listening to reviews of evacuations that have occurred over the past 10 years, some consistent patterns and lessons have emerged. As these lessons are learned and repeated, the knowledge gained from them appears to be becoming more deeply ingrained with later planning and practice. The following points summarize some of these lessons that may be taken from recent experience and applied elsewhere.
Evacuations are matters of life and death, not business as usual.
A common reaction of transportation agencies is to view evacuations as a large-scale version of a routine traffic event. While this approach has merit, it is not necessarily true.
Evacuations are emergencies on which the lives of many people may depend on the expeditious movement of traffic. As a result, the use of unconventional traffic management approaches such as contraflow, access and egress restrictions, and regional detours may be necessary. Some travel inconvenience to a non-evacuation traveler may be a small price to pay to evacuate threatened populations.
Evacuations are regional events.
Although the extent and effects of an evacuation depend on the size and advance warning associated with a particular threat, it is clear that most hurricane evacuations will have regional traffic effects. Over the past 30 years, population growth in hurricane-threatened coastal counties has grown enormously while the road capacity in the outlying areas of these regions has remained relatively unchanged. As a result, mass evacuations will rapidly overwhelm regional road networks, and their effects will be felt over hundreds of miles through downstream cities and counties and across state lines. Evacuation plans must account for these repercussions by evaluating the efficiency of entire corridors, including critical route intersection and merging points, even if they are far away and may be in another state.
Make attempts to spread demand spatially and temporally.
In Louisiana one of the most significant goals of the updated plan was to move more evacuees north rather than east. Although these may have not been the most direct routes to the ultimate destinations of the evacuees, these movements were critical because they allowed demand to spread more equally across the regional road network. Temporal spreading is often somewhat more difficult to accomplish. Many areas of attempted “phased” evacuations, in which the most vulnerable populations are encouraged to leave first, often are using geographic or political boundaries. However, evacuee movement also is heavily influenced by the actions of others and by recent history.
Maintain flexibility in response.
The conditions that precipitate evacuations are dynamic by nature. Hurricanes, in particular, have ever-changing forecast tracks and strengths. As a result, evacuation plans also should be flexible. The ability to implement or discontinue certain traffic management measures in response to changing threat, location and demand conditions is quite useful, particularly if they can be implemented rapidly.
There also are many other needs that cannot be covered in a single short article, some more obvious than others, including more effective plans for the use of transit for low-mobility evacuees, communication and coordination between agencies and with evacuees. On an operational level these include contracting for the early positioning of wrecker services and fuel suppliers and clearing shoulders. Others include provisions for clearing construction work zones so that they are usable during an evacuation and coordinating traffic-signal control to facilitate traffic movement in emergencies.
