中国机场延误模式动态时空异质性分析  

Analyzing dynamic spatiotemporal heterogeneity in Chinese airport delay patterns

作　　者：王中一 张恒才[1,2] 王培晓 於佳宁 康永伟 WANG Zhongyi;ZHANG Hengcai;WANG Peixiao;YU Jianing;KANG Yongwei(State Key Laboratory of Resources and Environmental Information Systems,Institute of Geographic Sciences and Natural Resources Research,CAS,Beijing 100101,China;College of Resources and Environment,University of Chinese Academy of Sciences,Beijing 100049,China;ESRI(Beijing)Software Research and Development Center,Beijing 100020,China)

机构地区：[1]中国科学院地理科学与资源研究所资源与环境信息系统国家重点实验室,北京100101 [2]中国科学院大学资源与环境学院,北京100049 [3]ESRI(北京)软件研发中心,北京100020

出　　处：《地理学报》2025年第2期490-502,共13页Acta Geographica Sinica

基　　金：中国科学院基础与交叉前沿科研先导专项(XDB0740100-02-05);国家博士后创新人才支持计划(BX20230360)。

摘　　要：延误模式反映了机场之间空间交互关系与特征,已有研究大多聚焦于静态或单一机场延误模式探索,忽视了机场延误的复杂非线性动态变化特点。本文提出机场延误模式动态时空异质性分析框架,通过延误网络构建、Gi*统计、Mann-Kendall趋势检测、网络节点影响力测度构建PageRankCube模型,对中国247个机场2021年10月约35万条航班记录捕捉进港延误、离港延误、飞行延误3种延误影响力网络的时空异质性模式。研究发现:(1)机场延误影响力的共性时空模式呈现出“东强西弱”,“南强北弱”,高置信度冷热点多中心聚集,较低置信度冷点分散分布的特点;(2)京津、长三角、珠三角与成都等大中型国际机场形成连续、增强的高延误时空影响力;东北地区、中部西部地区的中小型机场时空上呈现持续较低的延误影响力;(3)相较于离港延误网络,进港延误网络在时空分布上展现了更为显著的异质性特征,飞行延误与进港延误两种网络之间的时空异质性表现出较高的相似性;(4)增强热点型机场不仅具有强大的延误影响力,且易成为引发或传导大面积延误的关键环节;持续冷点型机场呈现出离散边缘化的结构特征,展现出“小世界”的网络特性。本文克服了传统静态延误模式分析的局限性,实现延误模式的动态时空异质性捕捉,为航空管控策略制定提供技术支撑。Flight delay reveals important spatial interaction patterns and could offer crucial insights into the unique characteristics of airports.Existing studies mainly focus on identifying static or individual airport delay patterns,but neglecting the complex nonlinear dynamics involved.To fill the gap,this study proposes a novel methodology to explore the variations in airport delays,considering both temporal changes and spatial differences.We develop a new PageRankCube model that adopts multiple technologies,such as delay network generation,G statistics,Mann-Kendall trend identification,and network node effect measurements.Three types of flight delay networks,including arrival delays,departure delays,and flight delays,are constructed to discover spatiotemporal variability patterns,utilizing a dataset of approximately 350,000 flight records from 247 airports in China.The results show that:(1)Airport delays have a greater impact in eastern and southern China,but less influence in western and northern China.Strong cold and hot spots are grouped in several locations,whereas weak cold spots are distributed.(2)Large international airports in Beijing,Tianjin,the Yangtze River Delta,the Pearl River Delta and Chengdu have formed a continuous and enhanced high delay spatial influence;small and medium-sized airports in the Northeast,Central and Western regions have shown a continuous low delay influence in time and space.(3)Compared with the departure delay network,the arrival delay network shows more significant heterogeneity in spatio-temporal distribution.The spatio-temporal heterogeneity between the flight delay and arrival delay networks shows higher similarity.(4)Enhanced hotspot airports can cause or transmit large-scale delays,while continuous cold spot airports have discrete and marginalized structural characteristics,resembling a"small world"network.This work addresses the constraints of conventional static delay pattern analysis,captures the changing spatial and temporal variations in delay patterns,and offers technic

关 键 词：机场延误 时序延误网络 PAGERANK 时空异质性 延误模式 中国 

分 类 号：V351[航空宇航科学与技术—人机与环境工程]