Hulkkonen et al. (2026) Europe’s extreme temperatures and rainfall in finer detail: strengths and limits of climate model downscaling

Identification

• Journal: npj Climate and Atmospheric Science
• Year: 2026
• Date: 2026-08-14
• Authors: Mira Hulkkonen, Akash Deshmukh, Tero Mielonen, Ian G. Brosnan, Taejin Park, Hugo Lee, Weile Wang, Bridget L. Thrasher, J. L. McCarty, Harri Kokkola, Anton Laakso
• DOI: 10.1038/s41612-026-01504-7

Research Groups

• Finnish Meteorological Institute, Yliopistonranta 8, Kuopio, 70210, Finland.
• NASA Ames Research Center, Moffett Field, California, 94035, USA.
• Bay Area Environmental Research Institute, Moffett Field, California, 94035, USA.
• Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, California, 91109, USA.

Short Summary

This study evaluates the accuracy and reliability of downscaled climate model outputs in Europe for extreme temperatures and rainfall. The results show that downscaling can add details to and increase the accuracy of modelled climate variables in Europe, but their reliability is variable- and region-dependent.

Objective

• To evaluate the representation of average temperature (tas), precipitation (pr), and their associated extremes (maximum of daily maximum temperature, TXx, and maximum precipitation accumulated over a 5-day period, Rx5day) in multiple global and downscaled climate model outputs.
• To assess the geographical dependence of accuracy and reliability of downscaled climate model outputs.

Study Configuration

• Spatial Scale: Europe-wide analysis with focus on regional patterns.
• Temporal Scale: Present day (1990-2019) and future (2071-2100) scenarios.

Methodology and Data

• Models used:
  • EURO-CORDEX (dynamically downscaled climate model output)
  • NEX-GDDP-CMIP6 (statistically downscaled and bias-corrected product)
  • GDPCIR (statistically downscaled and bias-corrected product)
  • CMIP5 and CMIP6 global climate models
• Data sources:
  • E-OBS observational benchmark
  • ERA5 reanalysis dataset
  • GMFD reanalysis dataset

Main Results

• The study found that downscaling can add details to and increase the accuracy of modelled climate variables in Europe, but their reliability is variable- and region-dependent.
• EURO-CORDEX exhibited a systematic negative bias for extreme temperature, particularly for lower maximum temperatures.
• NEX-GDDP-CMIP6 performed well for TXx, while GDPCIR ranked highest for tas and TXx.
• CMIP5 and CMIP6 global climate models outperformed the downscaled model outputs for average precipitation (pr) and Rx5day.

Contributions

• This study provides a comprehensive evaluation of the strengths and limitations of climate model downscaling in Europe, highlighting both the value and uncertainties associated with using downscaled climate model outputs.
• The results can inform decision-making processes related to climate change adaptation and mitigation efforts.

Funding

• Not specified.

Citation

@article{Hulkkonen2026Europes,
  author = {Hulkkonen, Mira and Deshmukh, Akash and Mielonen, Tero and Brosnan, Ian G. and Park, Taejin and Lee, Hugo and Wang, Weile and Thrasher, Bridget L. and McCarty, J. L. and Kokkola, Harri and Laakso, Anton},
  title = {Europe’s extreme temperatures and rainfall in finer detail: strengths and limits of climate model downscaling},
  journal = {npj Climate and Atmospheric Science},
  year = {2026},
  doi = {10.1038/s41612-026-01504-7},
  url = {https://doi.org/10.1038/s41612-026-01504-7}
}