2019 Volume 75 Issue 4 Pages 250-261
Damage from the Impending Tokyo Inland Earthquake is predicted to influence property demand and produce market disparities all over the city. The Tokyo government designates City Reconstruction Areas to prioritize reconstruction and redevelopment activities, but the majority of the city will likely fall under non-designated areas. Experiences from the Great Hanshin earthquake revealed that such non-designated areas generate many persistent empty lots. To alleviate the expected problem of empty lots, this research utilizes data from the Great Hanshin earthquake (approximately 1400 properties) to construct a Random Forest prediction model. The model is able to predict the presence or absence of land transactions of damaged property five years after the earthquake, at an accuracy rate of 81%. Further analysis of the model revealed two measures to promote land transactions: 1) property merging, and 2) allowing zoning reclassification from “commercial use” to “residential”. Impacts of the two measures were tested with data from Tokyo, and it is estimated that a combination of both measures is predicted to decrease empty lots by 30% after the Tokyo Inland Earthquake.