Research on Identification of Urban Fringe Areas and Construction of Ecological Security Pattern: A Case Study of Beijing, China

Abstract

Urban fringe areas are ecological barriers for cities, and building an ecological security pattern in these areas can promote sustainable and high-quality development of the urban environment. Taking Beijing as an example, this paper accurately determines the scope of urban fringe areas by segmenting the grid with road network data, creating labeled data, and training the urban fringe areas determination model. Moreover, the “ecological sensitivity-ecological connectivity” method and the minimum calculative resistance model are utilized to construct a detailed ecological security pattern on urban fringe areas. The findings indicate: (1) The use of road network data to segment the grid overcomes the drawbacks of the traditional fishing net segmentation method and improves the accuracy and reliability. (2) The quantitative analysis method constructs scientific and reasonable labeling data, and the urban fringe areas determination model is verified by accuracy and works well. The fringe areas of Beijing are mainly located in Shunyi, Changping, Tongzhou, Daxing and Fangshan districts. (3) Ecological sensitivity in Beijing’s fringe areas is relatively dispersed. A total of 28 ecological sources were identified, with an area of 31.35km2. They are mainly distributed in the northwest side of urban fringe areas and adjacent to the mountainous areas. (4) There are 37 ecological corridors, densely distributed in the northern and southwestern parts of Beijing’s fringe areas, indicating two distinct partitioned states. The ecological security pattern of Beijing’s fringe areas requires further improvement. These findings provide critical insights for sustainable urban ecological development and the optimization of ecological security pattern.

Keywords: Road Network Segmentation, Quantitative Label Construction, Urban Fringe Areas Identification, Ecological Sensitivity, Ecological Security Pattern