气候公平性视角下城市洪涝风险的多维驱动机制
Climate Inequality to Urban Flooding Risks Driven by Multidimensional Factors
徐浩文
同济大学建筑与城市规划学院 博士研究生
周士奇（通信作者）
同济大学设计创意学院 助理教授，博士，zhoushiqi@tongji.edu.cn
耿汐雯
同济大学建筑与城市规划学院 硕士研究生
徐小东
同济大学上海自主智能无人系统科学中心 博士研究生
摘要: 随着极端降雨事件频发，城市洪涝已成为高密度城市面临的重大挑战。以粤港澳大湾区为研究对象，构建“危害—暴 露—脆弱性”框架，综合多源数据开展洪涝风险评估与资源公平性分析。采用贝叶斯优化的LightGBM-SHAP方法，揭示 关键致灾因子的变化机制；并通过Dagum基尼系数与洛伦兹曲线评估防洪安全资源的空间分配不均。结果显示：洪涝风 险呈现沿海向内陆递减格局，高风险区集中于核心城市与河网密集区；安全资源存在显著空间不平衡，新兴区域的规划 割裂与基础设施孤岛化加剧风险暴露；关键致灾因子随风险等级升高从地形—物理因素逐步转向水文—生态因素;不透 水地表率、植被覆盖率是影响洪涝总体风险的核心变量。融合风险机制解析与公平性量化评估，以期为气候适应性城市 规划提供科学支撑与政策依据。
Abstract: Urban flooding has emerged as a significant challenge for densely populated cities, particularly in the context of increasing extreme rainfall events. This study focuses on the Guangdong-Hong Kong-Macao Greater Bay Area, developing a “hazard-exposure- vulnerability” framework to assess flood risks and resource equity using a variety of data sources. By employing a Bayesian- optimized LightGBM-SHAP method, the study uncovers the mechanisms underlying key disaster-inducing factors. Additionally, the distribution of flood defense resources is evaluated for spatial inequities using the Dagum Gini coefficient and Lorenz curve. The findings indicate that flood risk decreases from coastal to inland areas, with high-risk zones concentrated in core cities and densely riverine regions. There is a pronounced spatial imbalance in the allocation of safety resources, with the fragmentation of planning and infrastructure isolation in emerging areas exacerbating risk exposure. Key disaster drivers shift from topographical and physical factors to hydrological and ecological factors as risk levels rise. Impervious surface rate and vegetation coverage are identified as critical variables influencing overall flood risk. This research integrates risk mechanism analysis with equity- based assessments, offering scientific support and policy recommendations for climate-adaptive urban planning.
关键词：城市洪涝；风险测度框架；可解释人工智能；空间公平性；适应性策略
Keyword: urban flooding; risk assessment framework; explainable artificial intelligence; spatial equality; adaptive strategies
中图分类号：TU984
文献标识码: A
United Nations. Goal 11: sustainable cities and communities[EB/OL]. (2015-01-20)[2025-02-23].
https://sdgs.un.org/goals/goal11.
ZHOU S, ZHANG D, WANG M, et al. Risk-
driven composition decoupling analysis for urban flooding prediction in high-density urban areas using Bayesian-Optimized LightGBM[J]. Journal of Cleaner Production, 2024, 457: 142286.
WANG M, FU X, ZHANG D, et al. Assessing urban flooding risk in response to climate change and urbanization based on shared socio-economic
pathways[J]. Science of the Total Environment, 2023, 880: 163470.
CRICHTON D. The risk triangle[J]. Natural Disaster Management, 1999, 102(3): 102-103.
LEI Y, CAO H, ZHOU X, et al. Impact of land use/
land cover changes on urban flooding: a case study
of the Greater Bay Area, China[J]. IEEE Journal of
Selected Topics in Applied Earth Observations and
Remote Sensing, 2024, 17: 13261-13275.
PARK K, LEE M H. The development and applica-
tion of the urban flood risk assessment model for reflecting upon urban planning elements[J]. Water, 2019, 11(5): 920.
BHATTACHARJEE S, KUMAR P, THAKUR P K, et al. Hydrodynamic modelling and vulnerability
analysis to assess flood risk in a dense Indian city
using geospatial techniques[J]. Natural Hazards, 2021, 105(2): 2117-2145.
SMILEY K T, NOY I, WEHNER M F, et al. Social inequalities in climate change-attributed impacts
of Hurricane Harvey[J]. Nature Communications,
2022, 13(1): 3418.
JUNG M W, HADDAD M A, GELDER B K. Examining heat inequity in a Brazilian metropolitan region[J]. Environment and Planning B: Urban Analytics and City Science, 2024, 51(1): 109-127.
ZHANG Q, WU Z, CAO Z, et al. How to develop
site-specific waterlogging mitigation strategies?
Understanding the spatial heterogeneous driving forces of urban waterlogging[J]. Journal of Cleaner
Production, 2023, 422: 138595.
SABER M, BOULMAIZ T, GUERMOUI M, et al. Examining LightGBM and CatBoost models for
wadi flash flood susceptibility prediction[J/OL].
Geocarto International, 2022[2025-02-23]. https://
www.tandfonline.com/doi/abs/10.1080/10106049.
2021.1974959.
ZHU Y, TIAN D, YAN F. Effectiveness of entropy weight method in decision‐making[J/OL].
2020[2025-02-23]. https://onlinelibrary.wiley.com/
doi/10.1155/2020/3564835.
RUDIN C. Stop explaining black box machine
learning models for high stakes decisions and use
interpretable models instead[J]. Nature Machine Intelligence, 2019, 1(5): 206-215.
DIKSHIT A, PRADHAN B. Interpretable and explainable AI (XAI) model for spatial drought prediction[J]. Science of the Total Environment, 2021, 801: 149797.
FANG X, ZHANG Y, XIANG Y, et al. A spatial model for coastal flood susceptibility assessment using the 2D-SPR method with complex network
theory: a case study of a reclamation island in Zhoushan, China[J]. Environmental Impact Assess-
ment Review, 2023, 98: 106953.
MOLNAR C, KÖNIG G, HERBINGER J, et al.
General pitfalls of model-agnostic interpretation
methods for machine learning models[C]//
HOLZINGER A, GOEBEL R, FONG R, et al. xxAI
- Beyond Explainable AI: International Workshop. Cham: Springer International Publishing, 2022: 39-68.
HUANG Y, HONG X, ZHENG Y, et al. Assess-
ment and optimization of spatial equity for urban parks: a case study in Nanjing, China[J]. Ecological
Indicators, 2024, 166: 112449.
XU X, ZHOU S, XU H, et al. Advancing urban hub planning: a bibliometric analysis of concepts,
effects evaluation, and spatial design[J]. Land Use
Policy, 2025, 152: 107507.
CÔTÉ J N, GERMAIN M, LEVAC E, et al. Vulnerability assessment of heat waves within a risk framework using artificial intelligence[J].
Science of the Total Environment, 2024, 912: 169355.
SODEN R, LALLEMANT D, KALIRAI M, et al. The importance of accounting for equity in disaster risk models[J]. Communications Earth & Environment, 2023, 4(1): 386.
WANG M, LI Y, YUAN H, et al. An XGBoost-
SHAP approach to quantifying morphological impact on urban flooding susceptibility[J]. Ecological
Indicators, 2023, 156: 111137.
WANG M, FU X, ZHANG D, et al. Urban flooding risk assessment in the rural-urban fringe based on a Bayesian classifier[J]. Sustainability, 2023, 15(7): 5740.
翟国方. 气候变化背景下规划视角的城市雨洪
灾害韧性应对：关键概念、基本思路和通用框架
[J]. 城市规划学刊，2024（1）：29-37.
ZHAI Guofang. Resilient planning responses to urban flood disasters in the context of climate
change: key concepts, fundamental ideas, and a comprehensive framework[J]. Urban Planning Forum, 2024(1): 29-37.
MARTIN A J F, CONWAY T M. Using the Gini Index to quantify urban green inequality: a systematic
review and recommended reporting standards[J].
Landscape and Urban Planning, 2025, 254: 105231.
DAGUM C. A new approach to the decomposition of the Gini income inequality ratio[C]//SLOTTJE D
J, RAJ B. Income inequality, poverty, and economic welfare. Heidelberg: Physica-Verlag HD, 1998: 47-63.
KE G, MENG Q, FINLEY T, et al. LightGBM: a highly efficient gradient boosting decision tree[C]//
GUYON I, VON LUXBURG U, BENGIO S, et al.
Advances in neural information processing systems
30. Red Hook: Curran Associates, Inc., 2017.
SHEN J, CHEN C, YANG M, et al. City size, population concentration and productivity: evidence
from China[J]. China & World Economy, 2019, 27(1): 110-131.
MARITAN C A, LEE G K. Resource allocation and strategy[J/OL]. Journal of Management, 2017[2025-02-23]. https://journals.sagepub.com/
doi/full/10.1177/0149206317729738.
JURKOVIĆ Ž, LOVOKOVIĆ D. Corridor
planning–sustainable planning?[J]. Sustainability, 2023, 15(21): 15502.
NAGAMANI K, MISHRA A K, MEER M S, et al. Understanding flash flooding in the Himalayan
Region: a case study[J]. Scientific Reports, 2024, 14(1): 7060.
