Application of Remote Sensing and GIS for Dynamic Information Analysis of Ecological Security

Satellite images are now used as the primary source of spatial information for land-use pattern analysis. In this paper, land-use information is extracted by the maximum likelihood classification (MLC) and neural network classification (NNC) method, which is a combinative methodology of automatic extraction and recognition of land-use in Beijing City from remote sensing data. The temporal and spatial dynamic trend of the ecological security in Northeast Beijing from 1996 to 2005, which is based on the landscape ecology theory and the stability of landscapes by using landscape ecological security index (ESI). The results show that land-use information obtained by the combinative methodology of automatic recognition and extraction is accurate, the landscape pattern and its change closely associated with the strong disturbance of human activities. The ESI can make quantitative evaluation for the ecological environment and its evolution direction. The ESI level put up an ascending tendency from Chaoyang District to Miyun District, and the area of the low ESI is enlarging gradually with the acceleration of the urbanization. This result would be useful for establishing better future management strategies for ESI in urban-rural transition zones.