Han et al. (2026) Refining Public DEMs for Urban Waterlogging Simulation via Vector–Raster Integration

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

Identification

• Journal: Remote Sensing
• Year: 2026
• Date: 2026-04-03
• Authors: Bo Han, Xiaoman Qi, Xiaotong Qi, Yuebin Wang
• DOI: 10.3390/rs18071080

Research Groups

[Information not explicitly provided in the given text.]

Short Summary

This study develops a novel semi-automated technique to refine public 30 m resolution Digital Elevation Models (DEMs) to 1 m resolution for urban areas. The method significantly improves the accuracy of urban inundation simulations by correcting road and waterway elevations, leading to better topological representation and more reliable flood depth predictions.

Objective

• To develop a novel technique to refine public 30 m resolution Digital Elevation Models (DEMs) to 1 m resolution for urban areas, thereby improving the accuracy of urban inundation simulations by accurately capturing depressional characteristics and water levels.

Study Configuration

• Spatial Scale: Urban areas, specifically Polk County in Florida, USA, and the Central Business District (CBD) in Beijing, China. DEM refinement from 30 m to 1 m resolution.
• Temporal Scale: The study focuses on refining static DEM data for improved flood simulation, addressing pre-flood critical states and simulating flood depths under different rainfall scenarios. A specific temporal duration for the study itself is not defined.

Methodology and Data

• Models used:
  • Semi-automated vector–raster integration workflow for DEM refinement.
  • Volume matching algorithm (for urban flood depth simulation).
• Data sources:
  • Public 30 m resolution Digital Elevation Model (DEM).
  • Vector data (for roads and waterways).
  • Official 1 m LiDAR DEM (for direct verification in Polk County).
  • Theoretical rainfall amounts (for indirect validation in Beijing).

Main Results

• The developed semi-automated vector–raster integration workflow successfully refines public 30 m DEMs to 1 m resolution for urban environments.
• In Polk County, direct verification against an official 1 m LiDAR DEM showed significant accuracy improvements: the mean error (ME) improved by approximately 9%, the root mean square error (RMSE) by approximately 20%, and the Standard Deviation (SD) by approximately 65% compared with previous methods.
• In Beijing, indirect validation using a volume matching algorithm demonstrated that the refined DEM significantly improved the topological accuracy of river channels and the reliability of simulated flood depths under various rainfall scenarios.
• The study analyzed two distinct types of water accumulation behavior patterns based on the refined DEM.

Contributions

• Development of an innovative semi-automated workflow that integrates public raster and vector data to refine low-resolution DEMs for urban areas.
• Establishment of a zero-flood-depth baseline by correcting road and waterway elevations, significantly improving the accuracy of urban inundation simulations.
• Construction of a highly precise water accumulation model by leveraging known attribute information from vector data.
• Significant improvement in the reliability of simulated flood depths and the topological accuracy of river channels in complex urban environments.

Funding

[Information not explicitly provided in the given text.]

Citation

@article{Han2026Refining,
  author = {Han, Bo and Qi, Xiaoman and Qi, Xiaotong and Wang, Yuebin},
  title = {Refining Public DEMs for Urban Waterlogging Simulation via Vector–Raster Integration},
  journal = {Remote Sensing},
  year = {2026},
  doi = {10.3390/rs18071080},
  url = {https://doi.org/10.3390/rs18071080}
}