Jiao et al. (2026) Multi attribute refined identification of flood-affected bodies based on multi-source data fusion

Identification

• Journal: Journal of Hydrology
• Year: 2026
• Date: 2026-02-08
• Authors: Yutie Jiao, Zongkun Li, Wei Ge, Meimei Wu, Bo Wang, Yong Zhang, P. A. H. J. M. van Gelder
• DOI: 10.1016/j.jhydrol.2026.135104

Research Groups

• School of Water Conservancy and Transportation, Zhengzhou University, Zhengzhou, PR China
• College of Civil Engineering and Architecture, Henan University of Technology, Zhengzhou, PR China
• State Key Laboratory of Hydraulics and Mountain River Engineering, Sichuan University, Chengdu, PR China
• Safety and Security Science Group (S3G), Faculty of Technology, Policy and Management, Delft University of Technology, Delft, The Netherlands

Short Summary

This study develops a multi-attribute diagnostic framework for flood-affected bodies by fusing multi-source data, addressing challenges in urban land function identification and dynamic population distribution characterization. It proposes an ensemble learning model for optimal urban land function identification and a human-land relationship matching method for high spatiotemporal resolution population mapping, demonstrating reliable support for comprehensive flood disaster assessment.

Objective

• To accurately and comprehensively identify the multi-attributes of flood-affected lands and populations, overcoming limitations in urban land function identification accuracy and dynamic population distribution characterization, to inform differentiated flood prevention and mitigation strategies.

Study Configuration

• Spatial Scale: Grid-based diagnostic framework; regional division for spatial downscaling of population data, enabling high-precision flood simulations.
• Temporal Scale: Dynamic population distribution characterization, addressing issues like nighttime distortion in Location-Based Service (LBS) data.

Methodology and Data

• Models used:
  • Ensemble learning model (for Urban Land Function (ULF) identification)
  • Hydraulic simulation
  • Geographic Information System (GIS) analysis
• Data sources:
  • Multi-source data fusion (general)
  • Location-based service (LBS) data (for dynamic population distribution)
  • Data traversal and multi-scale fusion (for ULF identification)

Main Results

• An ensemble learning model was successfully constructed for Urban Land Function (ULF) identification, providing reliable data for economic loss assessment and subsequent population spatial interpolation.
• A human-land relationship matching method, based on spatiotemporal behavioral laws, was proposed to reduce bias in dynamic population distribution data, achieving spatial downscaling and generating high spatiotemporal resolution population maps.
• A grid-based diagnostic framework was developed by coupling hydraulic simulation with GIS analysis, enabling multi-attribute diagnosis of disaster-bearing bodies including land function, population size, water depth, and spatial location.
• Case studies demonstrated that this framework provides reliable support for the accurate and comprehensive identification of flood disaster-bearing bodies.

Contributions

• Addresses the accuracy limitations of urban land function identification by developing an ensemble learning model through data traversal and multi-scale fusion, eliminating subjective selection uncertainty.
• Overcomes sampling bias and nighttime distortion in Location-Based Service (LBS) data for dynamic population distribution by proposing a novel human-land relationship matching method and achieving high spatiotemporal resolution spatial downscaling.
• Introduces a comprehensive grid-based diagnostic framework that integrates hydraulic simulation and GIS analysis for multi-attribute identification of flood disaster-bearing bodies (land function, population size, water depth, spatial location).
• Provides enhanced data support for economic loss assessment, population spatial interpolation, and analysis of population mobility's impact on flood risk.

Funding

• Not specified in the provided text.

Citation

@article{Jiao2026Multi,
  author = {Jiao, Yutie and Li, Zongkun and Ge, Wei and Wu, Meimei and Wang, Bo and Zhang, Yong and Gelder, P. A. H. J. M. van},
  title = {Multi attribute refined identification of flood-affected bodies based on multi-source data fusion},
  journal = {Journal of Hydrology},
  year = {2026},
  doi = {10.1016/j.jhydrol.2026.135104},
  url = {https://doi.org/10.1016/j.jhydrol.2026.135104}
}