DAMAGE INVENTORY ESTIMATION FOR LARGE SCALE EARTHQUAKE DISASTER ASSESSMENT USING NIGHT TIME CITY LIGHT
DAMAGE INVENTORY ESTIMATION FOR LARGE SCALE EARTHQUAKE DISASTER ASSESSMENT USING NIGHT TIME CITY LIGHT
复制标题
DOI:
--
复制
发表时间:
2002
期刊:
影响因子:
--
通讯作者:
M. Takashima;H. Hayashi
中科院分区:
文献类型:
--
作者:
M. Takashima;H. Hayashi
Traditional damage assessment for large-scale disaster has following two problems. First is there is no method to collect up-to-date information about damage inventory such as population and the number of buildings for large target area in fine spatial resolution, especially in developing country. Second is multi-jurisdictional nature of human settlement has been ignored. Human settlement tends to extend beyond administrative boundary. Assessment should be done by each human settlement cluster, not by administrative jurisdiction. To address these problems, we focused on city light distribution captured by DMSP (Defense Meteorological Satellite Program) satellite on daily basis as an index of human settlement across the boundary. At first, we developed an automated method to estimate stable light intensity which reflects the city light distribution by accumulating a large amount of DMSP imagery. Then, the methods to identify metropolitan area and to estimate the size of population and the number of buildings were developed by analyzing the relationship between the spatial distribution of city light and that of population and buildings with 1km grid based on the population census and the business census. It was clarified that spatial change of metropolitan area from 1995 to 2000 estimated by city light distribution reflects actual changes in the population and that the method we proposed can reliably estimate the population and the number of buildings with 1km grid within the municipalities with more than 10,000 population and more than 10,000 buildings. We applied these methods to assess the damage size due to the Tokai earthquake which is expected to occur near future. Which human settlement clusters can be especially impacted and which municipalities should closely cooperate each other for this event was clarified by aggregating the damage estimated in each 1km grid using the metropolitan area estimated by our method.