EXTRACTION FOOD DESERT AREAS WITH DETAILED HOUSEHOLD DATA ESTIMATED BY MERGING THE DIGITAL MAPS AND THE POPULATION CENSUS

摘要

Many kinds of data go digital to be available at all time (e.g. the national population census). Limitations of these existing statistical data are that aggregating units are spatially large and inconsistent between each statistical data. Many previous studies have difficulties in monitoring of the actual state of household behavior in detail because of these limitations. In the case of the national population census, the limitations can be solved if locations of each person can be monitored i.e. if we can disaggregate the census data. The disaggregation means to allocate the aggregated census data onto building polygons based on areas of each building and the housing statistics data etc. With this data, demographic changes can be grasped at scales of microscopic level like city blocks to the macroscopic like the whole area of a city or a prefecture. Furthermore, we can grasp population change of the city by considering moving rates, migration and age distribution of each household. It is expected that our new data will be used for various fields in the future. One of the examples of fields is the "food desert" problem. Using our data, food desert areas can be detected in detail. Using the disaggregate census data, as an example, "food desert" areas are detected in details. "Food desert" problem is one of the critical problem especially in Japan. Many previous studies have also tried to detect food desert areas. However almost all studies can monitor in limited areas. This study shows a method to develop detailed household composition data to extract food desert areas visually and quantitatively in detail to mitigate the food desert problem of Japan.