Huang et al. (2026) Decoding flood hazard responses to reticular river network changes using interpretable machine learning

Identification

• Journal: Journal of Hydrology
• Year: 2026
• Date: 2026-07-11
• Authors: Shanheng Huang, Peng Wang, Zulin Hua, Yueyang Dong, Rongxu Chen, Yawei Hu, Shimin Tian
• DOI: 10.1016/j.jhydrol.2026.136008

Research Groups

• Henan Key Laboratory of YB Ecological Protection and Restoration
• Key Laboratory of Integrated Regulation and Resource Development on Shallow Lakes, Ministry of Education, College of Environment, Hohai University
• College of Civil Engineering, Bengbu University
• Yellow River Institute of Hydraulic Research, YRCC

Short Summary

This study examines the impact of structural changes in the Lixia River network on flood hazard patterns from the 1960s to 2016 using a multi-index evaluation system and interpretable machine learning methods.

Objective

• Investigate how river network restructuring influences spatiotemporal flood hazard patterns in reticular networks of low-lying delta plains.

Study Configuration

• Spatial Scale: Local (Lixia River network) and regional (low-lying delta plain)
• Temporal Scale: 1960s to 2016

Methodology and Data

• Models used: Hydrological-hydrodynamic modeling, interpretable machine learning methods
• Data sources: Historical data on river network structure, hydrological observations, satellite imagery, and reanalysis datasets

Main Results

• Flood hazards increased by 2.17%–5.01% from the 1960s to the 1980s, then declined but remained above 1960s baseline levels by 2016.
• Spatially, hazards concentrated in north-central regions with poor hydraulic connectivity, with centroids shifting westward by approximately 9.2 km.
• Network structural indices eccentricity and closeness centrality emerged as critical drivers comparable in importance to land use factors.

Contributions

• This study provides quantitative guidance for structural rehabilitation in reticular river networks of low-lying plain regions, highlighting the importance of network structure in flood hazard distribution.

Funding

• Henan Key Laboratory of YB Ecological Protection and Restoration
• National Natural Science Foundation of China (Grant No. 52179201)
• Yellow River Institute of Hydraulic Research, YRCC

Citation

@article{Huang2026Decoding,
  author = {Huang, Shanheng and Wang, Peng and Hua, Zulin and Dong, Yueyang and Chen, Rongxu and Hu, Yawei and Tian, Shimin},
  title = {Decoding flood hazard responses to reticular river network changes using interpretable machine learning},
  journal = {Journal of Hydrology},
  year = {2026},
  doi = {10.1016/j.jhydrol.2026.136008},
  url = {https://doi.org/10.1016/j.jhydrol.2026.136008}
}