Group 4a: Spatial Analysis on a Global Scale
Spatio-temporal Analysis of Landcover with GIS and Remote Sensing
Members of the Group
| OKABE, Atsuyuki | Professor, Department of Urban Engineering, University of Tokyo | ||
| Leader | |||
|
| |||
| SAKURAI, Yukio | Professor, Graduate School for Humanity and Sociology, University of Tokyo | ||
| Field survey and data collection | |||
| SHIBASAKI, Ryosuke | Associate Professor, University of Tokyo | ||
| Data conversion | |||
| SADAHIRO, Yukio | Lecturer, Research Center for Advanced Science and Technology, University of Tokyo | ||
| Data analysis | |||
Research Plan
The objective of this research project is to analyze landcover change in Islamic areas. In conventional methods, spatial data are collected through field survey and paper maps. Collecting those data, however, is usually very costly, and the collected data often contains various spatial and attribute errors. To overcome this shortcomming, we use GIS with remote sensing data which provide spatially precice data at a relatively low cost. Using this system, we analyze the temporal change in landcover patterns in Islamic areas.The study region consists of three subregions, two located in Islamic areas, one in a non-Islamic area. Each region is a square whose sides measure approximately one hundred kilometers. We obtain remote sensing data of three levels of resolution for each region, that is, 8 kilometers, 30 meters, and 3meters, to estimate the landcover of the regions. The obtained data is stored and manipulated in GIS, and then converted into the raster data of landcovers.
The landcover change is analyzed in three steps. First, we develop a method for describing the topological relationship between spatial patterns, and apply it to the patterns of landcover change. Second, introducing statistical methods, we test the significance of the landcover change. Third, by comparing two or more spatial patterns of landcovers, we estimate the factors affecting landcovers and analyze them. These three methods are applied to the three regions studied in this project, which will enable us to describe and analyze the landcover change in this region.
In the first year of the project, we collected the spatial data for analysis, and reviewed existing studies dealing with remote sensing, image analysis, and spatial statistics. We then obtained NOAA remote sensing data and converted it into the landcover data. In the second and third years, we are developing statistical methods for analyzing temporal change in landcover and apply them to the data obtained in the first year.