城市边缘区识别及其生态用地景观格局优化研究—

文章信息/Info

Title:: Research on the Identification of Urban Fringe Areas and Optimization of Ecological Land Landscape Patterns: A Case Study of Suzhou City

作者:: 田雅丝; 陈家乐; 施荟文; 钱万溪; 陆雨婷; 田雅丝，苏州大学建筑学院副教授，博士；陈家乐，苏州大学建筑学院硕士研究生；施荟文，苏州大学建筑学院硕士研究生；钱万溪，苏州大学建筑学院硕士研究生；陆雨婷，苏州大学建筑学院硕士研究生。

Keywords:: urban fringe areas; spatial range recognition; multi source geographic data; deep learning; landscape patterns; Suzhou City

摘要:: 城市边缘区空间范围的合理识别是生态用地管理和景观格局优化的前提。研究选取POI数据、夜间灯光数据、土地利用数据等多源地理数据，基于深度神经网络模型构建城市边缘区识别方法，以苏州市为例进行实证研究，依据识别结果对边缘区生态用地景观格局进行分析并提出优化策略。结果表明：苏州市核心区面积1 380.14 km2，占比15.90%，边缘区面积3 589.02 km2，占比41.50%，乡村面积1 017.34 km2，占比11.80%，整体呈现核心区—边缘区—乡村的梯度分布趋势；边缘区的生态用地景观破碎化程度较高，整体连通性较差；为实现边缘区生态用地景观格局优化的目的，针对边缘区过渡性、缓冲性和动态性特征提出了相应的优化策略。研究有助于认识城市边缘区生态用地景观格局特征，可为边缘区景观格局优化、生态用地规划提供科学依据。

Abstract:: The reasonable identification of spatial boundaries in urban fringe areas is essential for ecological land management and landscape pattern optimization. This study uses multi-source geographic data, including POI data, nighttime lighting data, and land use data, to develop a method for identifying urban fringe areas based on a deep neural network model, with Suzhou as a case study. The findings indicate that the core area of Suzhou City covers 1,380.14 km2 (15.90%), the fringe area 3,589.02 km2 (41.50%), and rural areas 1,017.34 km2 (11.80%), exhibiting a gradient distribution from core to fringe to rural. The ecological land in the fringe area is highly fragmented with poor overall connectivity. To optimize the ecological land landscape pattern of the fringe area, strategies are proposed that address its transitional, buffering, and dynamic characteristics. This research contributes to understanding the ecological landscape pattern of urban fringe areas and provides scientific basis for their optimization and ecological land planning.