Using time-serial geographic information such as olden topographic maps and olden aerial photos, the authors try to detect land cover and vegetation change or transition on Tama Hill area, Tokyo. The authors produce land cover data on Early Meiji Period using Jinsoku-sokuzu published by the Sanbou-honbu (the Japanese Olden Army), on just after World War Two using aerial photos taken by the US Army, and on the end of Showa Period using Vegetation Maps published by the Ministry of Environment. There are big differences of positioning accuracy and classification category on each geographic information. Vegetation Maps have low positioning accuracy compared with other geographic information. Jinsoku-sokuzu has low positioning accuracy in valley area. However, it is possible to detect not only development but also reforestation or succession. The authors can estimate that this method is useful to detect land cover or vegetation change over 90% area of study area, and letter information about vegetation on Jinsoku-sokuzu has important information of excellent vegetation condition on early Meiji period.
This paper shows the projection of population by age and sex with 500m mesh. We modify the cohort-change rate method in the following four points; interpolating secret data, calculating cohort-change rate using the data in the neighborhood of each mesh, revising the formula for computation of cohort-change rate and adjusting subtotal value. We projected 500m mesh-based population for 2005 using the Japanese 500m mesh-based population census data in 1995 and 2000 and examined accuracy of the projected data by comparing it with the observed population in 2005. As a result, the coefficient of correlation between the observed value and the estimated value is over 0.99. This shows that the result is capable of projecting population at an acceptable level.