都市構造と震災の様相

Abstract

The damage caused by an earthquakes does not relate only to the conditions of the natural environment. After the Southern Hyogo Prefectural Earthquake of 1995, many natural scientists addressed the spatial features of the damage, which formed a “belt” between the coastline and the mountains. But the damage was also caused by human and social conditions, related to the urban structure. This kind of situation which is affected by many kinds of natural and social conditions in each region, is studied using a geographic information system (GIS). Analyzing urban earthquake disasters using GIS may also be useful to simulate potential damage for disaster preventation in other cities. The final purpose of this research therefore is to develop such a method of GIS. The most appropriate spatial unit to analyze earthquake damage is the town district. Architects have already started to analyze the unit of individual building but this approach has some problems: The data and work volumes are huge and a large GIS is needed, so that we cannot develop such a brief system for use in other areas easily; Precise data on each buildings are private information and thus not easy to access; And spatial location is not crucial to the analysis of architectural structures. Consequently, the unit of a town district is a more convenient approach from the viewpoint of urban structure or urban planning. It can also include community data and so contribute to urban planning directly, for example, regarding the location of open space. In the case of Higashinada Ward in the eastern part of Kobe City, which recorded the largest number of persons killed by the earthquake, a “belt” of damaged buildings was found and studied by many natural scientists. Some assumed that a new hidden fault line existed along the belt. But the urban structure of Kobe City itself comprises a “belt” structure. Thus, before searching for such a fault line, we, need to analyze the relationship between urban structure and _disasters using data sources on the damage. There are maps of the distribution of damaged buildings, as recorded by City Planning Institute of Japan and Architectural Institute of Japan, and data from newspapers on the location of fatalities. We can also use general data sources regarding urbanization, for example, maps, aerophotos, and census data, because we need to be able to apply the same system in other cities. As shown in Fig.l, the damage is different even inside the “belt.” Generally, the collapse rate of building was extremely high in the built-up area in 1948, which suvived bombing during World War II. There were also some damaged districts near the coast outside the “belt, ” because of the same reason. Most areas along the coast were bombed and burned in World War II, so the buildings were rebuilt and relatively newer there, and were not severely damaged. In some old districts, many buildings were rebuilt after World War II for example, in the district along R. 43 which is the main industrial road between Osaka and Kobe. At the foot of the mountains in the north of this area, few old buildings were damaged sumed to be due to good topological conditions and the higher social class of the residents. In the eastern part of this area, many districts that were urbanized after World War II were severely damaged, because here were built many low-cost, cheap wooden apartments for young workers as shown in Fig. 2. Figure 3 shows that high death-rate districts correspond to the severely damaged districts shown in Fig. 1, and that not all the districts where many elderly people lived (Fig. 4) had a high death rate of the elderly, due to the superior conditions of the buildings. GIS analysis is very effective in studying the spatial factors of disasters in urban areas where complicated elements exist.