Supply-Demand risk assessment of urban flood resilience from the perspective of the ecosystem services: A case study in Nanjing, China

• A framework for assessing supply–demand risk of urban flood resilience. • The framework identifies flood-risk levels across different urban areas. • The machine learning method SOM network is employed for subdistrict clustering. • Both the supply and demand of urban flood resilience exhibit marked heterogeneity. • Supply-demand equilibrium spatial structure: central city—subcenter—functional zone. Urban flooding has become a pressing challenge for many countries and regions. Meanwhile, ecosystem-based disaster risk reduction approaches have been recognized as a sustainable and effective strategy for managing flood risks. This study designed a conceptual framework for assessing the supply–demand risk of urban flood resilience (UFR) from the perspective of ecosystem services (ESs). Taking the city of Nanjing, China, as an example, the InVEST model and the multi-criteria comprehensive evaluation method were employed to quantify the supply of UFR provided by natural ecosystems and the demand for UFR from socio-economic systems. Additionally, based on UFR supply–demand evaluation indicators calculated for each subdistrict, the Self-Organizing Map (SOM) was used to cluster the subdistricts. Finally, UFR supply–demand matching was conducted on the subdistrict clusters, and different flood-risk levels were identified based on the supply–demand ratio. The results showed that high flood-risk subdistricts are mainly concentrated in central urban area, low flood-risk subdistricts are primarily in urban periphery, and subdistricts in urban–rural transitional zones exhibit medium flood risk. Statistical analysis revealed that this zonal pattern is closely related to land use types and the distribution of social resources. Therefore, this study provides a scientific basis for developing management strategies of urban flood prevention from the perspective of ESs.