摘要
数智技术是提升城市暴雨态势感知能力的重要手段,从物理-信息-社会三元空间视角出发,通过fsQCA(fuzzy-set Qualitative Comparative Analysis)和LDA(Latent Dirichlet Allocation)的混合使用,解决数智手段赋能城市暴雨感知的路径问题,并探讨不同组态路径下的城市受灾特征。结果表明,实现高赋能水平存在3种模式:物理-信息-社会三元空间均衡型赋能、物理-社会空间主导的三元空间型赋能和物理-社会空间交互的二元空间型赋能。同时,一元空间的低赋能模式难以实现灾害的全面感知,验证了三元空间组态交叉的必要性。此外,LDA主题模型分析显示,不同的数智赋能模式匹配了不同受灾类型的城市,二元空间型赋能路径更适合致灾因子敏感型城市(如广州等),物理-社会空间主导的三元空间型赋能模式更匹配孕灾环境敏感型城市(如西安等),三元空间均衡型赋能模式更适合承灾体敏感型城市(如北京、上海等)。
Against the backdrop of rapid urbanization and escalating risks posed by extreme rainstorms,the complexity of urban hydrological systems and limitations of fragmented data-driven approaches underscore the necessity of constructing integrated frameworks to enhance rainstorm situation awareness.Traditional methodologies typically rely on isolated physical monitoring,digital modeling,or social response mechanisms and fail to address the interdependencies among physical infrastructure,informational technologies,and social systems.This study aims to deepen our understanding of how digital and intelligent technologies can be configured across a physical-informational-social ternary space to achieve robust urban rainstorm governance by identifying context-specific empowerment paths and their applicability to diverse urban typologies.Guided by the theoretical framework of the physical-informational-social ternary space,this study employs a mixed-method approach combining fuzzy-set Qualitative Comparative Analysis(fsQCA)and Latent Dirichlet Allocation(LDA)modeling to investigate the pathways through which digital and intelligent tools empower urban rainstorm perception and to explore the disaster-affected characteristics of cities under different configurational paths.By tracking 35 typical Chinese cities,the fsQCA analysis reveals three differentiated empowerment configurations:(1)Balanced ternary space empowerment(G1),which achieves high-efficiency empowerment through threedimensional collaboration among physical space data integration(including real-time sensor networks for hydrological monitoring),informational space intelligent analysis(including machine-learning-based risk prediction models),and social space emergency response(including interagency coordination systems),relying on dynamic interactions across the three domains.(2)Physically–socially dominant ternary space empowerment(G2):Grounded in core conditions of multisource data integration(combining meteorological,topographical,and citizen-generated data)and high disaster perception efficiency,this configuration incorporates peripheral conditions of server-side intelligence(including cloud-based data analytics)and user-side participation(including mobile application-driven hazard reporting),emphasizing data diversity and user-centric empowerment.(3)Physically–socially interactive binary space empowerment(G3):Empowerment is realized through the binary coupling of multisource data integration and high perception efficiency as the core conditions,prioritizing the technical synergy between physical monitoring and informational processing.Concurrently,a single-dimensional,low-empowerment configuration,which relies on isolated spatial data or technologies,is found to be insufficient for comprehensive disaster perception,thus empirically validating the necessity of ternary space configurational intersections.LDA topic modeling further demonstrates that different digital-intelligent empowerment patterns align with distinct disaster-sensitive city types:G3 suits hazard-sensitive cities(including Guangzhou),G2 matches vulnerable cities(including Xi'an),and G1 benefits exposure-sensitive megacities(including Beijing and Shanghai).Theoretical contributions of this study include constructing a"ternary space for urban rainstorm situation awareness"framework,which systematically analyzes the effects of digital-intelligent empowerment through the coupling mechanism of real-time physical space perception,intelligent informational space processing,and optimized social space decision-making—thereby transcending the limitations of traditional technological determinism.Methodologically,the research overcomes the constraints of single-method approaches by retaining fsQCA's strength in causal necessity analysis and integrating LDA's capability for semantic theme identification,forming a complete explanatory chain of"causal mechanisms-adaptive paths-type characteristics."At a practical level,this study proposes differentiated implementation strategies that provide both theoretical foundations and practical guidance for the digital and intelligent enhancement of urban rainstorm situation awareness.
作者
闫绪娴
王俊丽
温烜
刘杨
Yan Xuxian;Wang Junli;Wen Xuan;and Liu Yang(School of Management Science&Engineering,Shanxi University of Finance and Economics,Taiyuan 030006,China;School of Economics,Central University for Nationalities,Beijing 100081,China)
出处
《热带地理》
北大核心
2025年第9期1644-1656,共13页
Tropical Geography
基金
山西省2024年度研究生科研创新项目“三元空间视角下数智赋能城市暴雨态势感知的内在机理及匹配路径”(2024KY487)
国家社科学基金一般项目“我国城市暴雨内涝灾害形成机理、韧性评估及防治对策研究”(20BGL260)
山西省研究生教育创新计划支持“优秀研究生导师团队”(2024TD25)。
关键词
三元空间
数智赋能
城市暴雨态势感知
fsQCA
LDA
ternary space
digital-intelligent empowerment
urban rainstorm situation awareness
fuzzy-set qualitative comparative analysis
Latent Dirichlet Allocation
作者简介
闫绪娴(1978-),女,山西朔州人,博士,教授,主要研究方向为应急管理与韧性城市,(E-mail)yanxux@163.com;通信作者:王俊丽(1994-),女,山西晋中人,博士研究生,主要研究方向为应急管理与韧性城市,(E-mail)wjl18435176014@163.com。
