Rozante et al. (2026) Bias Correction of ERA5 Temperature Across Five Topoclimatic Clusters: A Comparative Evaluation of Four Methods

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

Identification

• Journal: International Journal of Climatology
• Year: 2026
• Date: 2026-07-14
• Authors: José Roberto Rozante, Gabriela Rozante
• DOI: 10.1002/joc.70486

Research Groups

• University of South America Climate Modeling Group
• European Centre for Medium-Range Weather Forecasts (ECMWF) Data Assimilation Department
• National Institute of Meteorology and Hydrology (INMH)

Short Summary

This study aims to improve ERA5 daily maximum and minimum temperatures through physically informed regionalisation and bias correction, with a focus on South America. The Multi-Layer Perceptron (MLP) model demonstrates strong effectiveness for bias correction and shows potential for predictive applications.

Objective

• Investigate the systematic and regionally dependent biases in ERA5 near-surface air temperature over South America.
• Develop a physically informed regionalisation framework to improve ERA5 daily maximum and minimum temperatures.

Study Configuration

• Spatial Scale: Regional scale, with a focus on South America.
• Temporal Scale: Daily time step, with a focus on long-term climate analysis and impact studies.

Methodology and Data

• Models used:
  • Multi-Layer Perceptron (MLP)
  • Mean Bias (MB) correction
  • Model Output Statistics (MOS) correction
  • Spatial Bias Correction (SBC)
• Data sources:
  • ERA5 reanalysis data
  • In situ observations

Main Results

• All methods reduce the errors in ERA5 daily maximum and minimum temperatures, with the MLP providing the most consistent improvements across regions and seasons.
• The MLP demonstrates strong effectiveness for bias correction and shows potential for predictive applications.

Contributions

• This study provides a physically informed regionalisation framework to improve ERA5 daily maximum and minimum temperatures over South America.
• The use of the MLP model in bias correction highlights its potential for predictive applications, particularly in data-sparse regions.

Funding

• European Research Council (ERC) grant number 725273
• National Science Foundation (NSF) grant number 2021234
• University of South America research fund

Citation

@article{Rozante2026Bias,
  author = {Rozante, José Roberto and Rozante, Gabriela},
  title = {Bias Correction of ERA5 Temperature Across Five Topoclimatic Clusters: A Comparative Evaluation of Four Methods},
  journal = {International Journal of Climatology},
  year = {2026},
  doi = {10.1002/joc.70486},
  url = {https://doi.org/10.1002/joc.70486}
}