Dou et al. (2026) Analysis of the performance of a virtual gauge-based method in hydrological modeling of basins with no precipitation stations

Identification

• Journal: Scientific Reports
• Year: 2026
• Date: 2026-03-04
• Authors: Yanhong Dou, Xiangning Liu, Xiao Liu, Min Xie, Ronghua Liu
• DOI: 10.1038/s41598-026-39531-2

Research Groups

• State Key Laboratory of Water Cycle and Water Security, China Institute of Water Resources and Hydropower Research, Beijing 100038, China
• Research Center on Flood & Drought Disaster Prevention and Reduction of the Ministry of Water Resources, Beijing 100038, China

Short Summary

This study evaluates the hydrological performance of the virtual gauge-based method (VG) for flood forecasting in basins without precipitation stations, demonstrating that VG-driven simulations achieved up to approximately 50% higher flood volume prediction accuracy and superior flood simulation capabilities compared to traditional methods.

Objective

• To evaluate the performance of the virtual gauge-based method (VG) in estimating areal rainfall and its hydrological performance for flood forecasting in basins completely devoid of precipitation stations.

Study Configuration

• Spatial Scale: Xiaoergou Basin, covering an area of 16,761 square kilometers, characterized by no internal precipitation stations but with adjacent stations.
• Temporal Scale: Daily hydrological data for flood seasons (May to September) from 2010 to 2019 (10 years of flood season data).

Methodology and Data

• Models used:
  • Virtual Gauge-based method (VG) for multi-source rainfall data fusion.
  • Thiessen polygon method (control group for areal rainfall estimation).
  • Random Forest algorithm for rainfall estimation at virtual gauges.
  • Inverse Distance Weighting (IDW) for spatial interpolation.
  • Xin'anjiang hydrological model for flood simulation.
  • Particle Swarm Optimization (PSO) for hydrological model parameter calibration.
• Data sources:
  • Observed daily flow at the Xiaoergou hydrological station (basin outlet).
  • Observed daily rainfall from 22 precipitation stations located in basins adjacent to the study area.
  • Observed daily evapotranspiration from three evapotranspiration stations in the vicinity.
  • Six daily real-time/short-latency multi-source precipitation products (0.1° spatial resolution):
    • GSMaP-N (Japan Aerospace Exploration Agency)
    • GSMaP-GN (Japan Aerospace Exploration Agency)
    • IMERG-E (NASA)
    • IMERG-L (NASA)
    • GRAPES (China Meteorological Administration)
    • ECMWF simulation and analysis products

Main Results

• The VG method adaptively adjusts the number, locations, and rainfall values of virtual gauges based on the spatial characteristics of rainfall fields, dynamically optimizing the weighting of precipitation products.
• Compared to the Thiessen polygon method, VG-driven hydrological simulations achieved significantly higher flood simulation accuracy.
• During the calibration period, the comprehensive accuracy metric (RMD) for VG was 0.88, compared to 0.61 for the Thiessen method, with flood volume prediction accuracy (RFWmean) improving by approximately 53% (12.99% for VG vs. 27.64% for Thiessen).
• In the validation period, VG maintained superior performance with an RMD of 0.85 compared to 0.72 for the Thiessen method, and the mean absolute error in time to peak (EPTmean) was notably lower (1.5 days for VG vs. 3.3 days for Thiessen).
• VG-driven simulations demonstrated a higher pass rate for flood event simulation, indicating greater reliability of hydrological performance.
• VG enhances flood simulation accuracy by improving the spatial distribution of rainfall estimated from sparse gauges through the dynamic construction of virtual gauges, effectively supplementing missed rainfall centers or adjusting rainfall magnitudes.

Contributions

• This study addresses a critical research gap by evaluating the hydrological performance of the virtual gauge-based method (VG) specifically in basins with no precipitation stations, extending previous work that focused on regions with sparsely distributed gauges.
• It demonstrates the superior capability of the VG method in providing more accurate rainfall estimates and flood simulations in data-scarce regions.
• The findings offer practical insights into the feasibility and benefits of applying the VG method for flood forecasting in ungauged or data-limited basins.

Funding

• The National Key Research and Development Program of China, grant number 2023YFC3006700
• Ningbo Municipal Water Resources Science and Technology Program Project, grant number NSKA202507

Citation

@article{Dou2026Analysis,
  author = {Dou, Yanhong and Liu, Xiangning and Liu, Xiao and Xie, Min and Liu, Ronghua},
  title = {Analysis of the performance of a virtual gauge-based method in hydrological modeling of basins with no precipitation stations},
  journal = {Scientific Reports},
  year = {2026},
  doi = {10.1038/s41598-026-39531-2},
  url = {https://doi.org/10.1038/s41598-026-39531-2}
}