Liu et al. (2026) The sensitivity of WRF extreme precipitation simulations over the Henan Province to different topography and reanalysis data

Identification

• Journal: Scientific Reports
• Year: 2026
• Date: 2026-01-06
• Authors: Songnan Liu, Aifang Su
• DOI: 10.1038/s41598-025-34545-8

Research Groups

• Henan Key Laboratory of Agrometeorological Support and Applied Technique, CMA, Zhengzhou
• Henan Meteorological Observatory, Zhengzhou

Short Summary

This study investigates the sensitivity of WRF extreme precipitation simulations over Henan Province to different topography data, reanalysis initial/boundary conditions, and model initial times. It found that WRF simulations are more sensitive to initial and boundary conditions than topography, with ERA IBC generally outperforming FNL IBC, and the impact of SRTM topography varying significantly depending on the IBC.

Objective

• To assess the sensitivity of WRF extreme precipitation simulations over Henan Province to different topography data (USGS, SRTM), initial and boundary conditions (FNL, ERA), and model initial times.

Study Configuration

• Spatial Scale: Henan Province, China (elevation 50 m to 2000 m). WRF nested domains at 27 km, 9 km, and 3 km horizontal resolutions. Reanalysis data at 1° x 1° (FNL) and 0.25° x 0.25° (ERA). Topography data at approximately 1 km (USGS) and 90 m (SRTM).
• Temporal Scale: Extreme precipitation event from 29 July to 31 July 2023. Simulations covered 72-hour and 24-hour accumulated precipitation. Model initial times were 0000 UTC on 27 July, 28 July, and 29 July, each with a 6-hour spin-up.

Methodology and Data

• Models used: Weather Research and Forecasting (WRF) model (version 4, implied by reference [29]).
  • Physical parameterizations: Thompson microphysical scheme; Kain-Fristch cumulus convective scheme (turned off in the innermost domain); Dudhia shortwave radiation; Rapid Radiative Transfer Model (RRTM) longwave radiation; Yonsei University boundary layer scheme; Noah Land surface model.
• Data sources:
  • Observation: Hourly observed precipitation data from 2752 stations in Henan Province (China Meteorological Administration).
  • Reanalysis (Initial and Boundary Conditions - IBC):
    • Fifth generation European Centre for Medium-Range Weather Forecasts reanalysis data (ERA) - 0.25° x 0.25° spatial resolution, 6-hourly frequency.
    • Final Operational Global Analysis data (FNL) from the National Center for Environmental Prediction (NCEP) - 1° x 1° spatial resolution, 6-hourly frequency.
  • Topography:
    • United States Geological Survey (USGS) - 30 arc-seconds (approximately 1 km) resolution (WRF default).
    • Shuttle Radar Topography Mission (SRTM) - 90 m resolution.

Main Results

• All WRF experiments generally captured the accumulated precipitation patterns, but significant differences were observed in the location, intensity, and timing of precipitation.
• The WRF model demonstrated greater sensitivity to Initial and Boundary Conditions (IBC) than to topography data.
• ERA IBC experiments generally showed superior performance in simulating various precipitation levels and the temporal distribution of precipitation compared to FNL IBC experiments.
• FNL IBC experiments initialized on 29 July better replicated the spatial distribution of precipitation, with FNL0729new achieving the highest Pearson correlation coefficients (CC) for both 24-hour (0.78) and 72-hour (0.84) accumulated precipitation.
• The use of SRTM topography improved performance (higher CC, lower Root Mean Square Error - RMSE) in most FNL IBC experiments for 24-hour accumulated precipitation.
• Conversely, for ERA IBC experiments initialized on 28 July, updating to SRTM topography led to a decline in the simulation performance for spatial precipitation distribution.
• The choice of initial time significantly influenced results, with models initialized at 0000 UTC on 29 July generally simulating precipitation amounts better.
• Experiments tended to underestimate precipitation maxima, although the location of maximum precipitation was relatively close to observations in some cases.

Contributions

• This study is the first to encompass a large ensemble of simulations to investigate the combined impacts of initial and boundary data sources and topography on extreme precipitation simulations.
• It provides crucial insights into how the combination of different topography data and IBC affects rainfall prediction in the WRF model over the Henan Province.
• The findings contribute to improving forecast skills for heavy rainfall events in the study region by highlighting the complex interplay of model configurations.
• Demonstrates that the impact of high-resolution topography data (SRTM) is not uniformly positive and can vary significantly depending on the initial and boundary conditions used.

Funding

• National Natural Science Foundation of China (Grant 42330610)
• National Key R&D program of China (2023YFC3209301)

Citation

@article{Liu2026sensitivity,
  author = {Liu, Songnan and Su, Aifang},
  title = {The sensitivity of WRF extreme precipitation simulations over the Henan Province to different topography and reanalysis data},
  journal = {Scientific Reports},
  year = {2026},
  doi = {10.1038/s41598-025-34545-8},
  url = {https://doi.org/10.1038/s41598-025-34545-8}
}