Waleed et al. (2026) A high-resolution global flood susceptibility dataset derived from multi-source earth observation and geospatial data

Identification

• Journal: Scientific Data
• Year: 2026
• Date: 2026-09-23
• Authors: Mirza Waleed, Muhammad Sajjad, Sami G. Al‐Ghamdi, Meng Gao
• DOI: 10.1038/s41597-026-08338-1

Research Groups

• Department of Geography, Hong Kong Baptist University, Hong Kong, Hong Kong SAR, China.
• Environmental Science and Engineering Program, Biological and Environmental Science and Engineering Division, King Abdullah University of Science and Technology (KAUST), Thuwal, 23955-6900, Saudi Arabia.
• Department of Geography and Resource Management, The Chinese University of Hong Kong, Hong Kong, Hong Kong SAR, China.

Short Summary

This paper presents the Global Flood Susceptibility Map (GFSM v1), a high-resolution global flood susceptibility dataset derived from multi-source earth observation and geospatial data. The GFSM v1 offers a globally harmonized 30 m flood susceptibility dataset produced by incorporating multi-source flood conditioning factors accounting for topographic, hydrological, meteorological, and anthropogenic characteristics within a machine learning framework.

Objective

• To develop a high-resolution global flood susceptibility map that can be used to identify areas prone to flooding worldwide.
• To provide a globally consistent and harmonized flood susceptibility dataset at 30 m resolution.

Study Configuration

• Spatial Scale: Global, with a spatial resolution of 30 meters.
• Temporal Scale: The study uses historical data from 2014 to 2024 for training the model.

Methodology and Data

• Models used: Gradient Boosting Machine (XGBoost) algorithm was used for modeling flood susceptibility.
• Data sources: Multi-source earth observation and geospatial data, including satellite imagery, digital elevation models, and global precipitation measurement datasets.

Main Results

• The GFSM v1 dataset achieved high accuracy in distinguishing flooded vs non-flooded locations within each region (median AUC ~0.95 and F1-score ~0.89 across all 192 units).
• The consistency of performance across diverse environments underscores the robustness of the chosen flood conditioning factors and modeling approach.
• The GFSM v1 dataset provides a globally harmonized 30 m flood susceptibility dataset that can be used to identify areas prone to flooding worldwide.

Contributions

• This study addresses the gap in high-resolution global flood susceptibility mapping by providing a globally consistent and harmonized flood susceptibility dataset at 30 m resolution.
• The GFSM v1 dataset offers a valuable resource for flood risk management, climate change adaptation, and disaster risk reduction efforts worldwide.

Funding

• This research was funded by King Abdullah University of Science and Technology (KAUST) and the Hong Kong Baptist University.

Citation

@article{Waleed2026highresolution,
  author = {Waleed, Mirza and Sajjad, Muhammad and Al‐Ghamdi, Sami G. and Gao, Meng},
  title = {A high-resolution global flood susceptibility dataset derived from multi-source earth observation and geospatial data},
  journal = {Scientific Data},
  year = {2026},
  doi = {10.1038/s41597-026-08338-1},
  url = {https://doi.org/10.1038/s41597-026-08338-1}
}