The spatial distribution of land price is indicative of the urban structure of a region. However, since most information on land price is in the form of point data and not surface data, a certain process must be established for the creation of a land price map. The final objective of the current study is to create accurate land price maps of three metropolitan areas in Japan: Tokyo, Nagoya and Kyoto-Osaka-Kobe metropolitan areas. This study uses the officially assessed land prices provided by the Japanese government and data from the land price surveys conducted by prefectural governments. First, the study presents empirical analysis of land price function of each metropolitan area by employing trans-Gaussian kriging, which enables us to consider both spatial dependence and difference of price elasticity by price ranges. Then, the study creates land price maps. Finally, we compare the created land price maps and discuss the nature of the differences in detail.
