Urban flood risk assessment based on DBSCAN and K-means clustering algorithm

Abstract

Urban flood risk assessment plays a crucial role in disaster risk reduction and preparedness. It is essential to mitigate flood risks and establish a comprehensive analysis of factors influencing flood risk, as well as classify risk levels, in order to provide a clear model for risk assessment. This article aims to propose an efficient assessment method that can classify urban flood risk levels and assist cities in flood risk management, particularly in identifying high-risk areas. The study area chosen for this method is the municipal district of Fuzhou City, located in Fujian Province, China. The proposed method utilizes the Urban Flood Risk Assessment Index, which is developed based on the risk-vulnerability framework. It integrates the combinatorial empowerment method, Density-Based Spatial Clustering of Applications with Noise (DBSCAN), and the K-means algorithm to cluster the quantitative risk factors, enabling a comprehensive analysis of the risk results. The findings demonstrate that areas characterized by intense extreme rainfall, lower elevation, gradual slope, high runoff coefficient, and high population density tend to exhibit higher flood risk. Moreover, the dominant factors contributing to high risk in different regions vary. The results obtained from this method align well with the distribution of historical flood points, indicating the effectiveness of the risk map prepared using this approach. In comparison to the results obtained from the single clustering method and the TOPSIS method used in traditional risk assessment, the proposed method can successfully identify high-risk urban flood areas. Consequently, this method offers a valuable scientific basis for urban flood prevention and control planning.

Keywords:

• Urban flood
• clustering algorithm
• regional evaluation
• risk mapping
• risk management

Data availability statement

The data used or analyzed during this study are not publicly available due to intellectual property rights or privacy restrictions.

Disclosure statement

No potential competing interest was reported by the authors.

Additional information

Funding

We would like to thank the team that provided the datasets. We sincerely thank the tutors for their guidance and suggestions in the research process. This research was supported by the National Natural Science Foundation of China (Nos. 32071776 and 41571490); the Natural Science Foundation of Fujian Province, China (No. 2020J01465); and the China Postdoctoral Science Foundation (No. 2018M640597).