摘要
城市生态学研究通常需要明确的城市边界,但现有研究中使用不同像元大小的遥感影像划分城市边界,缺乏统一的城市制图单元,导致不同遥感数据源量化的城市边界缺乏可比性,限制了城市生态环境的分析与对比。因此,提出一种基于城市路网斑块的建成区提取方法,通过统一制图单元提升城市边界提取结果的可比性。该方法基于路网和高分辨率影像分割出路网斑块,结合多源遥感影像的信息提取城市建成区。结果表明,新方法显著修正了建成区提取结果的破碎化、边缘锯齿化和形状不规则问题,可有效避免因数据源不同导致的不可比性。此外,新方法能够将建成区提取的总体精度提升2%—4%。基于路网单元的建成区提取方法能够更加准确和标准化地提取城市边界,为城市生态学研究提供了技术支撑。
Urban ecology research plays a crucial role in understanding the intricate relationships between urban landscapes and their ecological effects.A fundamental requirement for this type of research is the establishment of clear and consistent urban boundaries.Previous studies commonly used remote sensing images to map these urban boundaries.Due to the varying pixel sizes of different types of images,significant discrepancies arise in how urban boundaries are delineated.This lack of a standardized mapping unit poses substantial challenges,as it diminishes the comparability of urban boundaries derived from different pixel sizes.Consequently,this inconsistency can affect analyses and comparative studies of urban ecological environments,undermining the robustness of findings in urban ecology.In response to these challenges,we propose a novel method for extracting built⁃up areas based on urban road networks.This innovative approach aims to enhance the comparability and accuracy of urban boundaries by establishing a unified mapping unit.The methodology begins with the segmentation of road networks using high⁃resolution imagery.This segmentation is then complemented by integrating information from multiple remote sensing data sources,which enables more precise identification of urban built⁃up areas.The outcomes of this research indicate that the new method substantially improves extraction results by addressing common issues associated with built⁃up area delineation,such as fragmentation,jagged edges,and irregular shapes.These improvements are critical as they enhance the visual and analytical clarity of urban boundaries,making them easier for researchers and urban planners to use.Moreover,by standardizing the mapping units,the method effectively mitigates non⁃comparability issues that often arise from using diverse pixel sizes.Importantly,the overall accuracy of built⁃up area extraction is found to increase by 2%to 4%compared to traditional methods.The road network⁃based extraction technique not only refines the process of delineating urban boundaries but also strengthens the foundations of urban ecology research.With more accurately defined urban boundaries,researchers can conduct analyses that yield valuable insights into the interactions between urbanization and ecological processes.This has profound implications for urban planning,as it allows for better⁃informed decision⁃making and resource allocation.In summary,the proposed method represents a significant advancement in urban ecology research,promoting greater consistency,accuracy,and comparability in urban boundary delineation.This innovation not only enhances our understanding of urban ecological patterns,processes,functions,and services but also supports sustainable urban management practices,ultimately contributing to the resilience and sustainability of cities in an era of rapid urbanization.
作者
李沐森
钱雨果
郭辰萌
周伟奇
韩立建
王伟民
LI Musen;QIAN Yuguo;GUO Chenmeng;ZHOU Weiqi;HAN Lijian;WANG Weimin(Sate Key Laboratory of Urban and Regional Ecology,Research Center for Eco-Environment,Chinese Academy of sciences,Beijing 100085,China;University of Chinese Academy of Sciences,Beijing 100049,China;Beijing Urban Ecosystem Research Station,Research center for Eco-Environmental Sciences,Beijing 100085,China;Shenzhen Environmental Monitoring Center,Shenzhen 518049,China;National Field Scientific Observation and Research Station of Regional Ecological Environment Change and Comprehensive Management in Guangdong Greater Bay Area,Shenzhen 518049,China)
出处
《生态学报》
CAS
CSCD
北大核心
2024年第24期10989-11003,共15页
Acta Ecologica Sinica
基金
国家重点研发计划(2022YFF1301101)
国家自然科学基金(42261144742,42271102)
国家野外站建设项目(SZDL2023000962)。
关键词
建成区边界
路网斑块
景观格局
多源数据
built⁃up area boundary
road network patch
landscape pattern
multi⁃source data
