Evaluation of historical simulations in the HCLIM43-ALADIN regional climate downscaling ensemble over Europe in the EURO-CORDEX-CMIP6 framework
Regional climate downscaling has historically been complementary to global climate models, providing higher-resolution data for local-scale studies and climate services. We use the regional climate model HCLIM43-ALADIN to downscale six CMIP6 global models in eight simulations on a 12.5-km resolution European domain. This is the HCLIM community contribution to the EURO-CORDEX-CMIP6 multi-model downscaling ensemble involving several regional climate models and soon to be available publicly via the Earth System Grid Federation. We document the HCLIM43-ALADIN ensemble and evaluate the simulated results during the historical time period 1991−2020 against the gridded station observations E-OBS and the reanalyses CERRA, CERRA-Land, and ERA5, focusing on near-surface temperature and precipitation. In winter in the global climate model-driven HCLIM simulations, a cold bias of about 3 °C and underestimated precipitation by about 25% are present in North Europe, and a warm bias of about 1 °C and overestimated precipitation of about 50% are present in Central and Southern Europe. In summer, a cold bias of about 1 °C and overestimated precipitation of about 25% are present in Northern Europe, and a warm bias of about 2 °C is present in Central and Southern Europe. Overestimated precipitation by about 200% is present in parts of Southern Europe, but this value is also amplified by the very low observed precipitation. Some of the temperature and precipitation biases are explainable by large-scale pressure field biases, causing a reduced westerly flow from the Atlantic. Positive temperature trends of about 0.5 °C per decade are present nearly everywhere and especially strong over Central and Eastern Europe (up to about 1.5 °C per decade in some regions and seasons). A comparison between the driving global models and downscaled simulations shows that the regional model tends to improve temperature biases but worsen precipitation biases. In our subjective choice of model ranking, the EC-Earth3-Veg-driven simulations are the best-performing, and the MIROC6-driven simulation is the worst-performing. Simulations of future climate scenarios until 2100 with HCLIM43-ALADIN will be analysed in a separate publication. The presented simulations can be used for decision-making and will contribute to planned multi-model EURO-CORDEX-CMIP6 analyses.
Authors
- Peter Kuma (ORCID: https://orcid.org/0000-0002-0910-8646)
- Fuxing Wang (ORCID: https://orcid.org/0000-0001-7582-2752)
- Bert van Ulft (ORCID: https://orcid.org/0000-0001-8653-2312)
- Ole B. Christensen (ORCID: https://orcid.org/0000-0002-9513-2588)
- Petter Lind (ORCID: https://orcid.org/0000-0002-2691-2361)
- Grigory Nikulin (ORCID: https://orcid.org/0000-0002-4226-8713)
- Andreas Dobler (ORCID: https://orcid.org/0000-0001-9490-0840)
- Fredrik Boberg (ORCID: https://orcid.org/0000-0002-2589-8422)
- Oskar Landgren (ORCID: https://orcid.org/0000-0002-6264-8502)
- Patrick Samuelsson (ORCID: https://orcid.org/0000-0003-4387-6232)
- Aitor Aldama Campino (ORCID: https://orcid.org/0000-0001-8453-4322)
- Juan Carlos Sánchez-Perrino (ORCID: https://orcid.org/0000-0003-4916-6709)
- Carolina Nilsson
- Joakim Löw
- Michael Sahlin
Institutions
- Agencia Estatal de Meteorología (ES)
- Swedish Meteorological and Hydrological Institute (SE)
- Royal Netherlands Meteorological Institute (NL)
- Norwegian Meteorological Institute (NO)
- Danish Meteorological Institute (DK)
- Bolin Centre for Climate Research (SE)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-09-17
- DOI
- https://doi.org/10.5281/zenodo.22813644
- Primary Topic
- Climate variability and models
- Type
- preprint