Using Multi-temporal RS Image and Landscape Metrics to Map Urbanization Process

摘要

Multi-temporal remote sensing imagery is widely used to map the urbanization-caused land use temporal dynamics, which mainly appears as built-up growth. Remote sensing integrated with landscape metrics is also used to quantitatively describe the landscape pattern of the urban area in recent literatures. In this study, we propose a grid-based framework to quantitively map the land development stage in the urban area. Remote sensing classification method is used to obtain thematic land use maps. Built-up growth is then extracted from the multi-temporal classification results by post-classification change detection. Landscape pattern, which is quantitatively described by landscape metrics, is derived from the thematic land use maps. A grid-based method is used to analyze the land development stage based on landscape pattern and built-up growth. Finally, the development stage and its temporal changes are mapped.