Proceedings Article | 14 July 2003
KEYWORDS: Landsat, Argon, Ecosystems, Image classification, Vegetation, Clouds, Remote sensing, Carbon, Roads, Earth observing sensors
The Xilin River Basin, Inner Mongolia, China is dominated by grassland ecosystems. Human activities (e.g., crop cultivation, livestock grazing, urban development, construction of infrastructure) in the basin have resulted in substantial changes in land cover, particularly since late 1980s. In this paper we used Landsat Thematic Mapper (TM) and Enhanced Thematic Mapper Plus (ETM+) data to study the land use/cover change in the Xilin River Basin, Inner Mongolia. High-quality (less than 10% cloud cover) TM/ETM+ images (WRS124-29 and WRS124-30) acquired on four different dates (July 31, 1987, August 11, 1991, August 27, 1997, and May 23, 2000) were collected for this study. The Xilin River Basin is within two Landsat images (WRS124-29 and WRS124-30), and therefore, we mosaic WRS124-29 and WRS124-30 images and then cookie-cut it for the Xilin River Basin after preprocessing of the images. Based on the vegetation and land use characteristics of the basin, a hierarchical land cover classification scheme was created, including identification of major grassland types (e.g., Stipa grandis, Leymus chinensis). Then we conducted image classification using both supervised and unsupervised classification methods. Ancillary data such as vegetation map, soil map, topography maps, field survey data and literature of the Xilin River Basin were used for interpretation and labeling of spectral clusters as well as accuracy assessment of land cover classification. Finally, the resultant classification results of those four dates were used to quantify the spatial extent of land use/cover change in the basin and to develop the transitional matrix of land use/cover changes since 1987. The results derived from this study could be linked to biogeochemical models for assessing the consequence of land use/cover change on carbon and nitrogen dynamics of terrestrial ecosystems in the basin.