RECOGNITION OF URBAN FRINGE AREA BASED ON REMOTE SENSING IMAGE: A CASE STUDY OF GUANGZHOU-FOSHAN METROPOLITAN AREA

摘要

Urban fringe area, which situates between the urban and rural area, is the most dynamic and vulnerable system during the urbanization process. It's significantly meaningful to offer an accurate boundary definition of the metropolitan area, which can lay a strong foundation for the research of its land cover change pattern, the urban planning and ultimately, the implementation of the urban integration scheme. Taken Guangzhou-Foshan as an example, this study has attempted to provide a reliable method in defining the urban fringe area. Based on the "Fringe Effect" theory in landscape ecology, this study has renewed the existing RS information entropy model for defining the urban fringe by embracing scale theory, cartography and urban geography. The inflexion points of information entropy are identified in three different directions (latitude, longitude and radius), and try to put forward an accurate urban fringe area of the Guangzhou-Foshan metropolitan area. Several conclusions can be drawn from the study. 1. The urban fringe of the Guangzhou-Foshan metropolitan area covers an area of 2031.21km~2, 30.09% of the total study area. 2. The variability of the land cover pattern in the urban fringe area is much higher that it is in the urban core area and non-urban (rural) area, indicating an intersecting area of the rural and urban landscape; 3.The employment of optimal scale RS information entropy and inflexion recognition is more effective and reliable to define the urban fringe area, and provides a renewed and stronger foundation for its further study.