Chen et al. (2025) Integrating machine learning with NSGA-Ⅱ to assess the synchronization effects of stormwater disaster hazard and green–blue infrastructure

Identification

• Journal: Journal of Hydrology
• Year: 2025
• Date: 2025-12-18
• Authors: Jiaxuan Chen, Sisi Wang, Pingping Luo, Chong‐Yu Xu, Hongyu Zhao
• DOI: 10.1016/j.jhydrol.2025.134777

Research Groups

• School of Environmental and Energy Engineering, Beijing University of Civil Engineering and Architecture, Beijing, China
• Key Laboratory of Historical and Cultural Heritage Protection, Inheritance and Spatial Planning in Shaanxi Province, Xi’an, China
• School of Water and Environment, Chang’An University, Xi’an, China
• Key Laboratory of Subsurface Hydrology and Ecological Effects in Arid Region, Chang’an University, Ministry of Education, Xi’an, China
• Shaanxi Province Innovation and Introduction Base for Discipline of Urban and Rural Water Security and Rural Revitalization in Arid Areas, Chang’an University, Xi’an, China
• Xi’an Monitoring, Modelling and Early Warning of Watershed Spatial Hydrology International Science and Technology Cooperation Base, Chang’an University, Xi’an, China
• Key Laboratory of Eco-Hydrology and Water Security in Arid and Semi-Arid Regions of the Ministry of Water Resources, Chang’an University, Xi’an, China
• Department of Geosciences, University of Oslo, Oslo, Norway
• School of Architecture and Urban Planning, Jilin Jianzhu University, Changchun, China

Short Summary

This study developed a framework integrating urban stormwater disaster hazard assessment, machine learning, and multi-objective optimization to investigate the synchronization effects between stormwater hazard distribution and green-blue infrastructure (GBI) allocation. It found that GBI deployment in high-hazard zones is increasingly cost-effective for mitigating flood risk and runoff, especially under extreme precipitation events.

Objective

• To investigate the synchronization effects between urban stormwater disaster hazard distribution and green–blue infrastructure (GBI) spatial allocation.
• To conduct multi-objective optimization for stormwater management by integrating stormwater disaster hazard assessment outcomes.

Study Configuration

• Spatial Scale: Yantai city, China.
• Temporal Scale: Simulated 120-minute rainfall events with return periods of 0.5, 1, 3, 10, and 30 years.

Methodology and Data

• Models used: Random Forest (machine learning), NSGA-II (multi-objective genetic algorithm).
• Data sources: Urban stormwater disaster hazard assessment outcomes, simulated rainfall event data.

Main Results

• Green infrastructure (GI) deployment in high-hazard areas achieved optimal stormwater management performance.
• Average runoff reduction rates ranged from 20.2 % (0.5-year return period) to 5.2 % (30-year return period).
• Coupled flood risk reduction rates ranged from 28.3 % (1-year return period) to 8.7 % (30-year return period).
• GI-only scenarios exhibited diminished effectiveness under longer return period events.
• Green-blue infrastructure (GBI) in high-hazard zones became increasingly cost-effective as the return period increased.
• Optimal weight ranges for GBI in high-hazard zones were 0.32–0.84 for 5-year return periods and 0.30–0.92 for 10-year return periods.

Contributions

• Proposes a novel framework for assessing the synergistic effects of green-gray-blue drainage systems by integrating Random Forest with NSGA-II.
• Fills a research gap by integrating stormwater disaster hazard assessment outcomes into multi-objective optimization for GBI allocation.
• Substantiates the critical synchronization between GBI allocation and stormwater disaster hazard spatial distribution, particularly under extreme precipitation conditions.

Funding

Not explicitly stated in the provided text.

Citation

@article{Chen2025Integrating,
  author = {Chen, Jiaxuan and Wang, Sisi and Luo, Pingping and Xu, Chong‐Yu and Zhao, Hongyu},
  title = {Integrating machine learning with NSGA-Ⅱ to assess the synchronization effects of stormwater disaster hazard and green–blue infrastructure},
  journal = {Journal of Hydrology},
  year = {2025},
  doi = {10.1016/j.jhydrol.2025.134777},
  url = {https://doi.org/10.1016/j.jhydrol.2025.134777}
}