Top>Research>Problem of commuters who have difficulty returning home after a large-scale disaster in the Tokyo metropolitan area


Shigeki Toriumi

Shigeki Toriumi [Profile]

Education Course

Problem of commuters who have difficulty returning home after a large-scale disaster in the Tokyo metropolitan area

Shigeki Toriumi
Assistant Professor of Department of Information and Systems Engineering, Faculty of Science and Engineering, Chuo University


The Tokyo metropolitan area experienced tremors of intensity level upper 5 during the Great East Japan Earthquake. As a result, almost all railways temporarily stopped service. Since the earthquake occurred on a weekday afternoon, a large number of commuting workers and students were present within offices and schools in Tokyo. The stoppage of railways made it impossible for such people to return home. The majority of railway operators have established speed regulations, rules for stopping operation and safety inspection methods based on the intensity and peak acceleration of earthquakes. The intensity of the Great East Japan Earthquake required railway employees to confirm safety by walking along the tracks. Therefore, it was predicted from immediately following the earthquake that significant time would be required before railways could resume service, even if there was no damage to railway facilities.

Approximately 8 million commuting workers and students use railways on a daily basis in the Tokyo metropolitan area (Figure 1). More than half of these commuters have a commute time of an hour or more. This means that they will have difficulty returning home if railway service is interrupted. Railways are a high-speed, large-volume form of transportation. There is no alternative form of transportation that can fulfill a similar function. Even if roads are not damaged, buses and taxes do not have sufficient transportation capacity. Therefore, the majority of commuters who are unable to return home will wait in facilities such as offices and stations until railway service resumes.

Even before the Great East Japan Earthquake, it had been pointed out that a large number of commuters would have trouble returning home in railway service stopped in the Tokyo metropolitan area. In 2005, a report was issued by the Expert Panel for Response to an earthquake centered directly under Tokyo, which was established by the Central Disaster-Preparedness Committee of the Cabinet Office. According to this report, if an earthquake caused service to stop for lines in the Tokyo metropolitan railway network, there would be approximately 3.9 million commuters within Tokyo who are unable to return home. This number increases to approximately 6.5 million commuters when including the three prefectures which comprise the Tokyo metropolitan area.

Figure 1: 8 million commuting workers/students in the Tokyo metropolitan area (left: origin station, right: destination station) Data Source: Metropolis Transportation Census (2005)

Railway operation on the day of the Great East Japan Earthquake

Immediately after the earthquake, almost all railways stopped their service. This includes JR East Japan, the Tokyo Metro, Toei Railways and private railways. At shortly after 6pm, JR East Japan announced that service for railways in the Tokyo metropolitan area would not be resumed that day. When explaining this decision, JR East Japan stated confusion would result if crowds of customer expecting service to resume gathered in stations only to find out that service would remain stopped.

At 8:40pm, the Tokyo Metro resumed service for limited portion of the Ginza Line and the Hanzomon Line. Once this information had been circulated, stations were mobbed by crowds of people who had been waiting for service to resume, people in a hurry to return home, and people who had been awaiting further developments at their place employment. Stations became so crowded that it was impossible to ensure the safety of passengers, so service on the Ginza Line was ultimately stopped again.

Afterwards, routes such as other lines of the Tokyo Metro, the Toei Subway, Odakyu Railways, Keio Railways, Tokyu Railways, Seibu Railways and Sotetsu Railways were successively opened. There were also some lines that continued service throughout the night. As a result, people living along these lines were able to return home by the morning following the earthquake, although it took a great amount of time. Furthermore, it also gradually alleviated crowding at major stations. The alleviation of crowding was also most likely affected by the information that operation of railways such as JR East, Tobu and Keikyu would not be resumed on that day, as well as people being guided to government shelters (Figure 2).

Figure 2: Resumption of service on the day of the earthquake (Green: lines which resumed service; Gray: Lines which remained stopped)

Estimation of commuters who experienced difficult returning home

How many commuters experienced difficulty returning home on the day of the Great East Japan Earthquake? It is difficult to ascertain an exact number, so I would like to make an estimation using data from a survey of daily movement by people. Although slightly old, I will use data from the Tokyo Metropolitan Person Trip Survey (hereinafter referred to as "PT data") which was conducted in 1998. PT data was a survey of individual behavior to assess transportation factors such as "when," "why" (for example, commuting to work/school, business, shopping, personal affairs, returning home) "how" (for example, train, walking, private car, etc.) and "from where to where." By using this PT data, it is possible to extract the number of people who were outside (present at a location (for example, place of employment); in route to a destination) other than their home at the time of the Great East Japan Earthquake. I used the departure time, arrival time at destination and transportation method from PT data in order to estimate the location of people in route outside of their homes at the time of the earthquake. I then created demand for returning home using the commuters' current location as the origin and their homes as the destination.

Next, I prepared a system composed of road networks and railway networks for railways which resumed service. I then selected a route which would reduce walking distance to a minimum for the previously created demand for returning home. When creating this model, I assumed that a walking distance of less than 10 kilometers made it possible for everyone to return home and that a distance of 20 kilometers or more made it difficult for everyone to return home. For all other cases, I assumed that the ratio of people with difficulty returning home would increase by 10% for every additional kilometer of walking distance. This assumption considers individual physical fitness and recognizes that the same walking distance may be possible or impossible for different individuals (the aforementioned Central Disaster-Preparedness Committee used the same method for estimations). Results of the estimation showed that approximately 2.81 million people had difficultly returning to their homes. When examining the location of those people at the time of the earthquake, it was clear that a large number of stranded commuters occurred in the Tokyo metropolitan area. It was also ascertained that those stranded commuters sought to reach destinations in eastern Saitama Prefecture and northwest Chiba Prefecture. Such a distribution was most likely caused by the limited resumption of railway service from the Tokyo metropolitan area to Kanagawa, Tama and western Saitama Prefecture after the earthquake (Figure 3).

Figure 3: Distribution of commuters who had difficulty returning home (left: location at the time of the earthquake, right: destination)

*In November 2011, the Cabinet Office announced an estimated of 5.15 million commuters who experienced difficulty in returning home from the Tokyo metropolitan area on the day of the earthquake. This estimate was based on a survey. However, this estimated did not considering crowding or the number of people who were able to ride on railway lines which had resumed service. Therefore, the result differs from my estimate. (My research is prone to underestimation. Accordingly, 2.81 million should be considered as the absolute minimum number.)

Expectations towards private railway businesses

As described above, a large number of commuting workers and students in the Tokyo metropolitan area depend on railways. Employing corporations and the government should enact the primary response to commuters stranded by a large-scale disaster. However, private railways business must also respond in order to reduce the number of stranded commuters as much as possible and to avoid confusion. Examples of response included providing relief goods and a place to wait in stations, as well as information regarding nearby emergency shelters. However, the most important response is to quickly resume operation.

Still, we must realize that there is danger in quickly resuming service. Railways in the Tokyo metropolitan area function as a network. If only a portion of the railways are operating, there is the possibility that a high concentration of riders will make it impossible for that portion to function. Indeed, after the Great East Japan Earthquake, the Ginza Line was mobbed with riders after resuming operation and was forced to stop service again due to crowding. A similar phenomenon can be observed during past earthquakes, with confusion resulting from crowds of riders concentrated on a portion of operation routes. Therefore, it is obvious that cooperation between railway businesses is important for resuming operation. In other words, computer simulation should be used to find a pattern for resuming service so as to minimize the number of stranded commuters and reduce the possibility of confusion. Based on simulation results, railway businesses should adjust the time and lines for resuming service. There is an urgent need to establish an organization and contact channel which will enable such cooperation and adjustment.

Individual response

If an earthquake occurs directly under the Tokyo metropolitan area, there will most likely be damage to railways facilities and a large-scale blackout. In such a situation, the period for resuming railway operation may be anywhere from several days to weeks or even months. Therefore, it is also necessary to consider individual response by each person. I would like to propose 3 such responses.

(1) Assess the distance that you are able to walk
After the Great East Japan Earthquake, there were people who became tired while walking home and gave up. Waiting in unfamiliar areas other than around your home or place of employment can cause uneasiness and led to secondary accidents. Therefore, it is important for each person to assess the distance that they are capable of walking and not to attempt walking a distance that exceeds their personal limit. Furthermore, if a disaster occurs when you are away from home, it is important to consider heading towards your place of employment if it is nearby, instead of making a vain effort to return home.

(2) Assess the route home from your place of employment
Not many workers or students who use trains for their daily commute know a route for walking home or have an accurate idea of the distance involved. Moreover, in the case of intermediate stations, almost no one knows a route for walking home or the distance. It is important to realize that service may be resumed for only a portion of railway line and to assess the route home and distances from major stations. Such steps are useful for preventing the aforementioned need to give up while attempting to walk home. I recommend checking your route home by accessing map information sites which can be viewed daily on your computer or mobile phone.

(3) Reviewing methods for confirming safety and timing for returning home
During disasters, communication channels are congested (crowded) by many people trying to confirm safety. As a result, such confirmation is delayed. People who have not confirmed their family's safety or the condition of their home tend to hurry home. However, you should make a promise with your family to confirm each other's safety by using disaster message services provided by communication companies. Furthermore, once you have confirmed the safety of your family, it is important to avoid vain attempts to return home. Instead, you should wait at your place of employment and take action one conditions have calmed down.


The Great East Japan Earthquake has clarified the problem of stranded commuters in the Tokyo metropolitan area. Experts are concerned that an earthquake will occur directly beneath Tokyo in the future. In such an event, it is predicted that lifelines will be damaged and fires will occur. The resulting confusion may exceed even that of the recent earthquake. Instead of simply relying on support from the local government and corporations, it is important that each person makes individual preparations (mental preparation is particularly important).

Shigeki Toriumi
Assistant Professor of Department of Information and Systems Engineering, Faculty of Science and Engineering, Chuo University
Born in Yokosuka City, Kanagawa Prefecture. Graduated from the Faculty of Science and Engineering, Chuo University in 1997. Completed the Master's Program in information engineering at the Graduate School of Science and Engineering, Chuo University in 1999. After employment at Canon Incorporated, completed the Doctoral Program in information engineering at the Graduate School of Science and Engineering, Chuo University in 2007. After serving as a researcher at the National Maritime Research Institute, assumed his current position in 2008. Affiliated with the Operations Research Society of Japan, the GIS Association of Japan, the Japan Society for Industrial and Applied Mathematics, the City Planning Institute of Japan and the Japan Society of Traffic Engineers. Currently conducts research on mathematical methods and practical methods for solving various problems in transportation systems such as railroads, shipping and airports, as well as closely related urban, rural and environmental problems.
Spatial Information Technology Laboratorynew window