Zheng et al. (2026) Urban Pluvial Flood Prediction in Huai’an City Based on a Transformer–GNN Fusion Model

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Water
• Year: 2026
• Date: 2026-09-05
• Authors: Xin Zheng, Yandong Tang, Xi Yu, Kaiwen Xue
• DOI: 10.3390/w18172206

Research Groups

• Department of Hydrology and Water Resources Engineering, University of Science and Technology of China
• Key Laboratory of Urban Hydrological Cycle and Aquatic Environment, Ministry of Education

Short Summary

This study develops a spatiotemporal prediction model that integrates a Transformer and graph neural network (GNN) to predict urban pluvial flooding. The results show clear accumulation, delayed response, and spatial heterogeneity in urban pluvial flooding.

Objective

• Investigate the temporal evolution of regional average flood depth and spatial differentiation of inundated grid cells at the municipal scale during heavy rainfall events.

Study Configuration

• Spatial Scale: Municipal scale (Huai'an City)
• Temporal Scale: Hourly time series data for a 24-hour period

Methodology and Data

• Models used: Transformer-GNN model with fusion attention mechanism
• Data sources: Multi-source data including hourly meteorological observations, terrain, land cover, drainage networks, and water-system data

Main Results

• Regional average flood depth follows a continuous process of low-level stability, sustained rise, rapid increase, delayed peak, slow recession at a high level, and rapid recession during heavy rainfall events.
• Inundated grid cells show a pattern of concentrated distribution in urban built-up areas, secondary distribution in county-level built-up areas, and scattered distribution in non-construction land.
• Spatial differentiation of medium- and high-grade inundated grid cells results from the combined effects of low-lying terrain, local relative depressions, and impervious surfaces in construction land.

Contributions

This study provides methodological support and decision references for urban flood risk identification, grid-based risk management, and emergency dispatch during extreme rainfall events. The developed spatiotemporal prediction model can be applied to other cities with similar urban morphology and climate conditions.

Funding

• National Natural Science Foundation of China (Grant No. 51879005)
• Key Research and Development Program of Jiangsu Province (Grant No. BE2019006)

Citation

@article{Zheng2026Urban,
  author = {Zheng, Xin and Tang, Yandong and Yu, Xi and Xue, Kaiwen},
  title = {Urban Pluvial Flood Prediction in Huai’an City Based on a Transformer–GNN Fusion Model},
  journal = {Water},
  year = {2026},
  doi = {10.3390/w18172206},
  url = {https://doi.org/10.3390/w18172206}
}