Japan is a disaster-prone country. In recent years disasters have become more severe and frequent, such as the 2011 earthquake off the Pacific coast of Tohoku, the 2016 Kumamoto earthquake, and the 2018 heavy rainfall in western Japan. In a disaster, schools and public halls become shelters, and many residents evacuate to these locations [1
]. Regarding these shelters, the Disaster Countermeasures Basic Law was amended in 2013, requiring the basic municipalities to designate shelters (Article 49-7 of the Disaster Countermeasures Basic Law), and consider evacuees in locations other than designated shelters. It stipulates (Article 86-7) that aspects of the living environment in shelters must be improve, such as the distribution of food, clothing, medicines and the provision of health care services (Article 86-6).
However, in the 2016 Kumamoto earthquake disaster, many evacuees evacuated outside the designated shelters, such as staying in the car, making it challenging to identify the evacuees and delaying support [3
]. Also, in the event of a large-scale disaster, local government officials cannot take the lead in managing shelters from the perspective of a lack of human resources. On the other hand, the government’s “Guidelines for Ensuring a Good Living Environment in shelters” [4
] recommends independent management by residents, but at present, there are few cases in which the discussion of division of roles in shelter management are thoroughly exhausted. Furthermore, if the department responsible for managing shelters and the department responsible for the health status of evacuees are separated within the local government, there is concern that health hazards may occur, such as delays in improving the shelter environment due to lack of cooperation. Based on this situation, it is necessary to review the designated shelters and determine the desired management entity and the health crisis management method for the victims.
Therefore, in this study, we focused on the status of prior arrangements regarding shelter management to examine the standard operation procedure (hereafter “SOP”) of a smooth shelter management system. The purpose is to understand designation of shelters (hereafter “DS”), operating entity and operation manager of shelters (hereafter “OEM”), and information sharing between the shelter and the disaster response headquarters/health and welfare department (hereafter “IS”). It is significant in that it provides suggestion for making advance preparation and a SOP of shelter management.
2. Literature Review
Many studies have clarified the actual situation of shelter operation from the viewpoint of a single municipality in each disaster and the case of a certain shelter [2
]. However, with a limited approach to a certain shelter, there are many parts due to the peculiarities of the area [7
], making it difficult to generalize. In addition, Ariyoshi et al. [8
] clarified the status of the formulation and utilization of shelter management manuals based on a mail survey conducted nationwide. For example, the author of the manual, the status of a revision, the type of standard version/facility version, storage location, etc.
In recent years, it has been envisioned that shelters will be operated together with residents and support staff in addition to local government officials in the disaster area. In order for various leaders to jointly carry out smooth shelter operations, it is necessary to decide which shelter will be used (designation status of shelters), who and how shelters are operated (operation system) and how to share shelter information (information sharing), but a simple survey on these subjects is not sufficient. Also, the shelter management guidelines [9
] of the Cabinet Office are composed of four perspectives, (Ⅰ) establishment of the shelter management system (normal time), (Ⅱ) operation of the shelter (after the disaster), (Ⅲ) response to needs, (Ⅳ) elimination of the shelter. Preparation is essential for smooth shelter operation, and there are a lot of tasks that must happen in regular times related to shelter operation. For example, the establishment of shelter operation system, the designation of a shelter, advance assumptions of initial concrete action, the establishment of a support system, measures for evacuees at home and people who have difficulty returning home. Also, the importance of information sharing such as the designation of shelters, the operation system of shelters, information management, and transmission is taken up as a checklist.
Therefore, in this paper, we first focus on DS, OEM, and IS, and organize past studies.
Firstly, during the disaster preparedness period, shelter location and evacuation routing operations are important. Researchers have proposed several models for shelter location and evacuation routing [10
]. Kilci et al. [11
] proposed a mixed integer linear programming-based methodology for selecting the location of temporary shelter sites and validated the model by using a city in Turkey as a case study. Bayram [12
] classified previous research on evacuation planning and management models based on traffic assignment models. Overall, researchers have proposed several models for shelter location and evacuation routing: first as separate and then, more recently, as combined [10
]. During the disaster response period, Haga et al. [13
] clarified the characteristics of shelters for each facility type such as local community centers and elementary and junior high schools, and Sasaki et al. [14
], investigated the actual use of temples as designated shelters. Also, since it is challenging to identify evacuees other than designated shelters such as undesignated shelters, necessary supplies and support may not be delivered [15
]. In Mashiki Town at the time of the 2016 Kumamoto Earthquake, there was no function to integrate information on undesignated shelters, and information integration was delayed, and there was a lack of awareness of the occurrence of evacuees other than designated shelters and their support [16
]. It was also clarified that the designated shelter had a function as a part of the support base in response to the request from the undesignated shelter. In addition, there was thought that it was necessary to dispatch local government officials to the designated shelter, and awareness that they could not respond due to the shortage of local government officials.
Next, the operating entity and operation manager of shelters will be described as examples of shelter operations in past disasters. Yamori [5
] showed an example of a shelter in which the operating system was gradually changed from time to time while utilizing volunteers under the strong regional leader in the Great Hanshin-Awaji earthquake. He also emphasized the importance of the evacuees themselves being involved in the operation for the evacuees’ independence. After the Niigata Chuetsu-Oki earthquake, the local community was living in an evacuated situation while operating the shelter independently [17
]. Kobayashi et al. [6
] clarified three groups of residents responsible for voluntary management: local leaders, community activity leaders, and circle activity participants. It was shown that inter-organizational cooperation based on the existing organization is necessary. In the Great East Japan earthquake, all teachers played a central role in the shelter located at the Ishinomaki City Hebita Elementary School. Also, the teachers took command of the operation of the shelter, and the principal made various decisions [2
]. The shelter operation manual in the “Miyagi Disaster Prevention Education Basic Policy” [18
] states, “Although the operation of shelters is the business of the municipalities, there are many cases where cooperation is required even on the side of providing facilities, so be familiar with the contents”. However, when the degree of damage was extensive and a wide area was damaged, such as the Great East Japan earthquake, the functions of municipalities were confused and the situation assumed by this manual was not reached.
Finally, regarding information sharing between other departments, one of the lessons learned from the Great East Japan earthquake was that it was not possible to grasp shelter and evacuees information, and to take effective measures based on appropriate situational awareness [19
]. Bando et al. [20
] stated that it is necessary to share information between the disaster prevention department of the local government and other departments/within each department, in order to uniformly grasp the shelter situation for realizing effective response. In addition, although information sharing within departments has improved, the current situation is that information from other departments is not being distributed across the board.
As described above, regarding DS, there are studies on the intention to use undesignated shelters that occur after a disaster. However, the study about undesignated shelters that the local government grasps in advance is not enough. Regarding OEM, the state of operation by various entities, and the elements necessary for the operation by a non-local government are clarified from actual individual cases. However, the national trend of cognitive sharing is not clear. Regarding IS, the necessity of information sharing has been shown, but few studies are focusing on shelters.
Based on the above, considering that preparation is vital for smooth shelter operation, there is novelty in that we focused on the status of three advance arrangements regarding shelter operation, DS, OEM, and IS. These results are useful in order to make a standard operation procedure of shelter management because they provide suggestions from the perspectives of designating shelters, examining the operating entity and operation manager, and establishing an information-sharing system. As a method of this research, we use a questionnaire survey of local governments’ disaster prevention departments nationwide.
This research was jointly conducted by the University of Tokyo Institute of Industrial Science and Tokushima Prefecture. The target of this study is the department in charge of disaster prevention in local governments nationwide. Table 1
shows the outline of the survey conducted. The number of valid responses was 420.
shows the population size by the municipality nationwide (national census, 2015 survey). According to this, it can be seen that 29.4% of all 1741 municipalities have a population size of less than 10,000, and 68.8% have a population of 50,000 or less. In contrast, as shown in Figure 2
, the population size by municipality and percentage of survey respondent municipalities nationwide show similar trends. It can be understood then that the results obtained in this paper reflect a national trend.
shows the rank of the staff member who submitted the response. Directors, chief clarks and chief examiners account for more than half of the respondents. The chief respondent is the most common answer, but it is also positioned as the contact person for this survey. It cannot be said that the chief has prepared the answer.
The disaster experience is defined as whether or not the Disaster Relief Act was applied. Here, the Disaster Relief Act’s application status announced by the Cabinet Office is targeted from 2009 to 2015 before this survey. As shown in Figure 4
, the number of responding municipalities for which the Disaster Relief Act was applied even once is 204 (about 49% of all responses, 204/413), and there are 136 responding municipalities applied once and one municipality applied it five times, respectively. In addition, the number of municipalities to which the Disaster Relief Act has been applied since 2015 is 154 (about 37% of all responses, 154/413). Comparing after 2009 and after 2015, the number of local governments that applied it one time wass higher after 2015 than after 2019. This is because the number of times is reduced to one when totaled after 2015.
shows the composition of the questionnaire. Most of them are selective, but some questions are descriptive. Selective answers is the format to has three or more answer categories for a question and has one or more answers selected from among them. On the other hand, the descriptive answer is the format where respondents are free to write. We use descriptive answers when asking about issues based on actual experience because it was assumed that the contents of the answers would be diverse.
The structure of this study is as follows: Section 4
shows the results of prior arrangements regarding shelter operations. The results of DS, OEM, and IS are shown. In addition, issues based on experience will be described for each business process of shelter management. Finally, in Section 5
, based on the clarified issues, the direction of the SOP of the shelter management system is considered from the viewpoint ofDS, OEM, and IS.
The purpose of this study is to clarify the local governments’ current situation of undesignated shelters, operating entity and operation manager of shelters and their recognition sharing, and the information sharing system between the shelter and the disaster response headquarters/health and welfare department, to create a SOP in Japan.
As a result, although there is a study by Ariyoshi et al. [8
] as a nationwide survey on shelter management, the following results were obtained in this paper compared to these related studies in the past.
Regarding the designation of shelters, it is necessary to regularly check the facilities and disaster safety of existing designated shelters. On the other hand, in undesignated shelters, safety in the event of a disaster has not been confirmed. There is a disparity in support between designated and undesignated shelters, so it is highly necessary to proceed with shelters’ designation.
Most local governments assume residents’ participation regarding the shelter operating entity and the operation manager. However, it can be said that the idea is not sufficiently known to the residents. It became clear that local government officials felt that many evacuees lacked self-help consciousness. In the future, it will be necessary to study how to describe it in regional disaster prevention plans and district disaster prevention plans and to promote public awareness and sharing with residents.
Regarding information sharing at shelters, about 1/4 of the local governments do not have a communication system with the disaster response headquarters. Also, the information sharing system with the health and welfare field has not progressed any further. It is necessary to establish clear rules and repeatedly examine the effectiveness of the rules, such as whether the defined information sharing means will function in a disaster.
In addition to the above three points, the items and issues to be considered were clarified by dividing them into the opening, acceptance, closing, and overall shelter management. In particular, the management and delivery of shelter keys, the response to residents, and tidying up due to closure are not described in the shelter management guidelines [9
] of the Cabinet Office, so it can be said that these should be added as a new perspective.
It is also essential to take measures against infectious diseases in the operation of shelters, such as the state of emergency being issued on 7 April 2020, following the spread of the new coronavirus. Even at present, the response to infectious diseases is stored in a database, but the content is insufficient, so we would like to enhance this.