CORDEX CORE Simulations

CORDEX Coordinated Output for Regional Evaluations (CORE)

High resolution regional climate information for the world


With the growing demand for high-resolution information about regional climate change and its impact all over the world, the WCRP CORDEX developed the CORDEX-COmmon Regional Experiment (CORE) Framework. CORDEX-CORE aims at contributing to the next IPCC report with a homogeneous dataset of high-resolution regional climate information and providing this information for all major inhabited areas of the world.

The main ideas of the CORDEX CORE framework are:

  • to provide an ensemble of high-resolution (at least 25 km) regional climate change information.
  • to provide a basis for assessments of future extreme events for all major inhabited regions of the world.
  • to use the ensemble for further analysis such as climate change impacts on the different global warming levels, e.g. +1.5 °C or +2.0 °C.
  • to be incrementally extended with further contributions by additional models/experiments
  • CORDEX CORE (~25km) is extending (not replacing) the CORDEX (~50km) regional climate information and provides a basis for further downscaling activities


RCMs currently contributing to the CORDEX CORE framework simulate at least nine domains (Figure 1) with horizontal resolution of 0.22° (about 25 km, see Figure 1a and 1b):

  • REMO model (contribution by GERICS)
  • RegCM model (, Italy, with the participation of Institute of Atmospheric Physics, Chinese Academy of Sciences (IAP/CAS), Oak Ridge National Laboratory, Tennessee, National Center for Atmospheric Research, NCAR, The Hong Kong University of Science and Technology)


  • evaluation: reanalysis, 1979 to 2017
  • historical, RCP2.6 and RCP8.5: GCMs, 1950 (1970) to 2100

Forcings: ERA-Interim re-analysis,   GCMs representing the range of low, medium, and high global equilibrium climate sensitivity (ECS) (in addition, there are three backup GCMs, as alternative forcing). The driving GCMs selected for the CORDEX-CORE ensemble offer a broad spread of ECS and were based on the suitability for the dynamical downscaling (McSweeney et al, 2015):

  • NCC-NORESM (low ECS, backup: GFDL-ESM2M)
  • MPI-ESM-LR/MPI-ESM-MR (medium ECS, backup: EC-Earth)
  • HadGEM-ES (high ECS, backup: MIROC-MIROC5)
CORDEX CORE model domains used in the simulations

Simulation matrix

CORDEX CORE Simulation Matrix

Data availability

The regional climate simulations of the CORDEX CORE activity are now available at the ESGF, look for the XXX-22 domains.

For other RegCM simulations not available at the ESGF, please find further information on data access here.

First results

Recent presentations at EGU 2020 showed first results using the CORDEX-CORE ensemble. Please refer to the presentations of the Session Regional climate modeling, including CORDEX

Assessing mean climate change signals in the global CORDEX-CORE ensemble

The complete presentation of a study, held at EGU 2020, by Claas Teichmann can be found here.

Mean annual temperature climate change signal at RCP2.6

Mean annual temperature climate change signal at RCP2.6

Ensemble mean temperature climate change signal (CCS): about +0.5 to +3.5 K (high CCS over North America and European domains).

Mean annual temperature climate change signal at RCP8.5

Mean annual temperature climate change signal at RCP8.5

Ensemble mean temperature climate change signal: about +0.5 to +4.5 K (about +1.0 K increase of CCS of the 2m temperature at the end of the century)

Climate indices projections based on the CORDEX-CORE, CMIP5 and CMIP6 ensemble

The complete presentation of a study held at EGU 2020 by Erika Coppola can be found here.

Extreme & Hazards - P99 validation (wet indicator)

The 99th percentile of precipitation (P99) as represented in observations and different ensembles is shown below.

The three ensembles show substantially different performances over the various domains

The three ensembles show substantially different performances over the various domains. In particular:

  • The CORDEX-CORE ensemble reproduces the location of the La Plata basin maximum, while it displaces the Amazon maximum slightly to the west compared to observations, with some underestimation in the eastern side of the basin.
  • The CMIP5 ensemble substantially underestimates the P99 throughout the continent and shows a misrepresentation of the spatial distribution.
  • The CMIP6 ensemble still underestimated the two maxima but it improves their spatial distribution compared to CMIP5.

Conclusion: the higher resolution of the RCMs may play a role in the improved reproduction of the extreme precipitation signal for both cases of complex topography and local land surface feedback mechanisms. This may also be the reason why the CMIP6 ensemble that has a higher resolution compared to CMIP5, is occasionally closer to the CORDEX-CORE ensemble and in general shows higher intensities.

Data Policy

We strongly encourage people using the data from the CORDEX-CORE database to contact the model data producers in order to give feedback on the model simulations, interact on the scientific studies and/or propose co-authorships.


For further information, please contact:

  • Claas DOT teichmann AT hzg DOT de
  • Coppolae AT ictp DOT it
  • Armelle DOT remedio AT hzg DOT de


Reference Papers for CORDEX-CORE evaluation:

Several papers about CORDEX-CORE climate projections are submitted and under review.