Land-cover Classification of China Based on Integrated Analysis of AVHRR Imagery and Geo-spatial Data

摘要

Information regarding the characteristics and spatial distribution of Earth's land-cover is critical to global environment research, and is increasingly required by global research community. In the past two decades, studies have been done in large-area land-cover mapping employing AVHRR imagery and geographic information in concert. However, such endeavors have seldom carried out in China. The purpose of this study was to generate land-cover classification map of China and evaluate the effectiveness of this approach. First, the study region was divided into nine sub-areas according to the climatic condition. Then, geo-spatial data including temperature, precipitation and elevation are resterized into image respectively, and then used as three bands to incorporate with remote sensing image layers to generate a composite image containing both geo-environmental and spectral information. Supervised classification was performed to generate the land-cover map of China. Although standard accuracy assessments were not done, a series of consistency check using available source of land-cover information were conducted. The result showed that the methodology used in this research is, in general, feasible for large-area land-cover mapping in China.