Land Use Change Analysis Using Regional Numerical Land Information 
Mitsunori Yoshimura 
Remote Sensing Technology Center of Japan 
Sumio Muraoka and Koji Imabayashi 
Ohba Co. LTD 
Commission IV, Working Group 1 
KEY WORDS: Geomatics/GIS, Land Use, Change Detection 
ABSTRACT: 
The regional numerical land information has constructed in Japanese three largest urbanized areas, Tokyo, Nagoya and Osaka 
and their suburbs. Its information is more effective for detail and microscopic investigation than the use of national land 
information. In this study, in order to discuss the potential of its land information, land use trends and change analyses were 
performed. The land use ratio, mixture and continuity with unit of 1 km mesh area was used for both analyses. As the results, 
land use ratio was proved to be used effectively for the regional land use trend investigation. The land use mixture and 
continuity were useful for detail land use change consideration. Furthermore land developing pattern and its transition could 
estimate by multi-temporal analysis. 
1.Introduction 
1.1 Regional Numerical Land Information 
Both national and regional numerical land informations 
are known to be published and used as the common land 
use data base in Japan. 
The national numerical land information has constructed 
as nationwide information data base and can be used 
effectively for the macroscopic land use, change 
investigation. 
On the other hand, the regional numerical land 
information has constructed in Japanese three largest 
urbanized areas, Tokyo, Nagoya and Osaka and their 
suburbs. Its information is more effective for detailed and 
microscopic investigation than the use of national land 
information. 
In this study, in order to discuss application fields using 
regional land information, the authors conducted the 
experimental study of investigations for data 
characteristics, regional land use trend and detailed land 
use change. 
1.2 Test Site and Used Data 
The test area of this study is west part of metropolitan 
Tokyo with 28kmX21km area. The used regional 
numerical land information consists of four temporal land 
use data with 10m mesh size. 
This information was generated based on '74 housing 
investigation as the first period, "79 as the second, '84 as 
the third and '89 as the fourth respectively. 
Further following four sub-test sites with 2kmX2km area 
were selected in order to execute our experimental study 
more detail. 
1) Area-A (existing urban area) 
2) Area-B (Housing development in agricultural area) 
3) Area-C (Development area by Governmental 
authorities) 
4) Area-D (Development area by private companies) 
2.Characteristics of Regional Numerical Land 
Information 
2.4 How to generate Regional Land 
Information ? 
Basically this regional numerical land use information 
generates or updates using base map(1/10,000), aerial 
photograph, resident map, change detection map 
(1/10,000). : 
Processing procedure is as follows: 
a) change detection by aerial photograph interpretation 
b) vectorization changed area 
c) coding land use number 
d) vector-raster conversion 
2.2 Verification/Comparison 
It is necessary to understand quality and characteristics of 
used land use data for discussing application fields. 
Every city, town and village have established own land use 
information by urban planning related projects. 
In order to verify used data, comparing of some land use 
data analysis results and existing statistical materials was 
carried out. 
Following two existing statistical materials used: 
1) Land use Tama city (1987) 
Area measurement based on 1/2500 land use map 
2) Land use Tokyo (1987) 
Area Measurement from 1/2500 Land Use Map 
2.3 Data Error and Noise 
Most of data error and noise are estimated to be caused on 
either a) change detection or d)vector-raster conversion. 
As the comparison results, following error or noise were 
identified: 
1) Noises at boundary of huge or line shape changed area. 
2) Misunderstanding about land use item 
3) Noises at connected area of maps 
4) Error depend on the generated date 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B4. Vienna 1996