Abstract
Accurate map work is critically important to determine distances of hibakusha(atomic bomb survivors) from the hypocenter, which is needed to evaluate radiation doses. However, the distances of hibakusha from the hypocenter were accurate for a 100-m mesh unit up to now because of the restrictions of map and space analysis technology. This study aims to undertake a more accurate evaluation by using a geographic information system(GIS). The outline of the procedure is as follows.
1)Selection of the base map. We selected a Hiroshima city urban planning map(scale:1/2500) made from 1925 to 1928 as a base map. To work with the GIS, the map was converted into a tiff file by a scanner.
2)House numbers for the map. This base map shows neither address information (house numbers) nor positional coordinates (latitude and longitude) but it is a survey map. Then, we added the house numbers from a Cadastre map made in around 1930.
3)Georeferencing. Georeferencing was undertaken using a GIS to give positional coordinates for the map made in step2 above. Georeferencing was used to select the control point (for example, buildings that have been exposed to radiation), to give the location information, and to adjust the distortion of the image.
4)Decisions on and digitizing of hibakusha locations by collation with ABS (database of atomic bomb survivors in Hiroshima) location information. The house polygon was made by searching on the map for a point (house) referring to the hibakusha location information(address) from the ABS and digitized geographic information. And, this house polygon was converted into the gravity point.We consider this gravity point as a hibakusha location.
5)Calculation of the distances of hibakusha locations from the hypocenter. The distance from a gravity point(step4 above) was calculated from the hypocenter by using Hawth Tools that was one of the GIS extension software.
As a result, we were able to confirm the effectiveness of this method though there were a few errors.
In this presentation, first of all, we will show details of the work. Next, we will report on the result up to the end of October 2009. Finally, we will try to examine the improvement from existing methods.
This work was supported by JSPS Grant-in-Aid for Challenging Exploratory Research(No.21652070).
