This study evaluated the convenience of post-disaster recovery housing complexes constructed in Japan's Miyagi Prefecture following the 2011 Tohoku Earthquake. We expected that five years on from the disaster, the evacuees would find life there more convenient. The study was targeted at five municipalities in the northern coastal area of Miyagi Prefecture that were most heavily struck by the earthquake-triggered tsunami: Kesennuma, Minami-Sanriku, Onagawa, Ishinomaki, and Higashi-Matsushima. Using geographical information system software, we conducted a network analysis of the coordinate data on the location of crucial infrastructure. Housing complexes within the defined service area of each infrastructure facility were assigned a score of 10, while those within two and three times the defined service area were assigned scores of 5.0 and 3.3, respectively. The convenience of the housing complexes was evaluated using these scores. In Kesennuma City, the post-disaster recovery housing complexes were shown to have better access to key infrastructure than the ordinary housing complexes, and in other three municipalities, the two types of complexes were shown to have comparable access. In Onagawa Town, however, the post-disaster recovery housing complexes were shown to have poorer access to key infrastructure than the ordinaly housing complexes.
The convenience of living environments is often evaluated considering the distance between the dwelling and
surrounding infrastructure (
The research focused on three cities and two towns in the northern coastal area of Miyagi Prefecture [Figure 1], Kesennuma, Minami-Sanriku, Onagawa, Ishinomaki, and Higashi-Matsushima, all of which were inundated by the tsunami that struck after the Tohoku Earthquake of 2011.
The data used in the study were the coordinates of the post-disaster recovery housing complexes [Figure 2], the ordinary housing complexes [Figure 2], and the infrastructure supporting daily life in December 2016. Post-disaster recovery housing complexes are the permanent public housing sites built after the disaster. On the other hand, ordinaly housing complexes are the permanent public housing sites built before the disaster. 2.2
The infrastructure evaluated in this study comprised the facilities listed below. The area assumed to be serviced by the facilities is given in parentheses. These service areas were defined considering the distance which could be accessed easily on foot.
1) Retail a. Supermarkets (1 km) b. Convenience stores (0.5 km) 2) Financial services
a. Post offices and banks (1 km)
3)Medical services
a. Clinics (emergency department: internal medicine, pediatrics, surgery) (1 km)
b. Clinics (no emergency department: others)(1 km)
The post-disaster recovery housing data were retrieved from the Miyagi Prefecture website (
After plotting the infrastructure sites on a map, we conducted a network analysis using the ArcGIS 10.2.2 software with the Network Analyst function (Esri Inc., Redlands, CA, USA). This software enables the true measurement of road distances, as well as the linear distances between two features. The software was run on a standard desktop PC.
We first created network datasets from simple road distances, neglecting bends and speed limit data. These network data sets were used to determine the service areas of the infrastructure, with individual features represented by polygons.
Based on the distance between the infrastructure and the area serviced, we then assigned a usability score to each housing complex. Post-disaster recovery housing complexes within the radius of a defined service area were assigned a score of 10, complexes within 2 of the defined service area were assigned a score of 5.0, and those within 3 of the service area were assigned a score of 3.3. The service area measured by the road distance for each facility has already been defined in section 2.3. For example, for supermarkets, the defined service area of the road distance is 1 km. These scores were used to assign a convenience rating to the housing complexes. 3
Convenience stores Supermarkets Post offices Banks Clinics(Urgent areas) Clinics(Others) Averages Based on these figure and tables, the findings were as follows:
In Kesennuma City, the convenience scores of the post–disaster recovery housing complexes were higher than those of the ordinary housing complexes. This means that the post-disaster complexes were more accessible to essential facilities than the ordinary housing complexes.
However, in Onagawa Town the post-disaster complexes achieved lower convenience scores than the ordinary housing complexes. This means that the post-disaster recovery housing complexes in Onagawa Town were less accessible to essential infrastructure than the ordinary housing complexes.
In other three municipalities, the convenience scores of the post-disaster complexes and ordinary housing complexes were almost the same, suggesting that they offered the same accessibility to essential facilities. 4
These results suggest that the daily life of evacuees still had not been convenient in Onagawa. These results were not in line with our prior expectations. For Kesennuma, the convenience of access to facilities might have been the first priority when deciding the location of the post-disaster housing complexes, whereas in the case of Onagawa, we assumed that the key consideration had been connection to the local community. We assumed that the criteria for the location of other three municipalities had been intermediate between those applied to the other two municipalities.