What is the Grand Ensemble?

The Max Planck Institute for Meteorology (MPI-M) Grand Ensemble (MPI-GE) is the largest ensemble of a single state-of-the-art comprehensive climate model currently available. The Grand Ensemble consists of five Large Ensembles of 100 simulations each, of the Max Planck Institute Earth System Model in the low resolution set up (MPI-ESM1.1), run with varying initial conditions. The five ensembles can be seen in the header image and are:

  • Historical (blue)
  • RCP2.6 (yellow)
  • RCP4.5 (green)
  • RCP8.5 (purple)
  • 1%CO2 (pink)

The Grand Ensemble also has a 2000 year preindustrial control simulation, run with 1850 conditions. 




How do I access and cite the MPI-GE data?

Atmospheric and ocean data (excluding biogeochemistry) is currently available. The biogeochemistry and land data is anticipated to be available in late February.

Data can be found here: esgf-data.dkrz.de/projects/mpi-ge/

To use the data the license must first be accepted and can be found here: https://esgf-data.dkrz.de/projects/esgf-dkrz/ 

Alternatively a direct download of any file will prompt the license agreement to appear. After the license is accepted the wget download will work. Please note that these scripts do not work when the safari browser is used. 

The left hand side of the webpage can be used to search for variables as can the bar at the top. For example to search for precipitation in the RCP8.5 scenario one can put 'pr AND RCP8.5' into the search bar. The most efficient way to download the whole ensemble is to search for the data that you want then click 'Add all displayed results to Data Cart'. From here go into the Data Cart and click 'Select All Datasets' then create a wget script for that download. To run the wget script the following command can be used: $ bash wget-xxx.sh -H 

The data will be available for 10 years from publication (end of 2028).

For scientists who wish to use the data before it is publicly available, we welcome collaboration with scientists based at the Max Planck Institute. 

Please cite the following publications when using the data.

The overview publication (Maher et al.) is currently submitted to JAMES and can be found in submitted format here: Maher_etal_submitted


If your publication requires all references to be accepted please currently cite the following publications instead: 

  • Stevens, B. (2015), Rethinking the Lower Bound on Aerosol Radiative Forcing, J Climate, 28(12), 4794–4819, doi:10.1175/JCLI-D-14-00656.1
  • Bittner, M., H. Schmidt, C. Timmreck, and F. Sienz (2016), Using a large ensemble of simulations to assess the Northern Hemisphere stratospheric dynamical response to tropical volcanic eruptions and its uncertainty, Geophys. Res. Lett., 43(17), 9324–9332, doi: 10.1002/2016GL070587.


Please send an email to grandensemble@we dont want spammpimet.mpg.de if you publish a paper using the data, find any errors or bugs in the data or would like to request data that is not currently available. 

The data is currently available on the native grid. We recommend CDO as a tool for regridding, selecting subregions, selecting time periods and much more. It is fast, efficient and can be used in scripting languages (bash, python) to easily loop over ensemble members. CDO can be found here: https://code.mpimet.mpg.de/projects/cdo/.





A list of publications using the MPI-GE >>>




Collaborative projects involving MPI

The SMILE (Single Model Initial-condition Large Ensemble) group was initiated together with colleagues from LMU Munich, University of Reading, and NCAR to provide a forum for exchange and discussion for the large ensemble community.



A special issue on large ensembles​ based on the EGU session on large ensembles has been proposed for ​Earth System Dynamics​. T​his issue is open to submissions which exploit the new opportunities offered by large ensembles and explore how a combined analysis of the existing large ensembles can advance our knowledge in different fields. The special issue particularly invites submissions that use new methods to investigate these topics.

Tentative contributions, including a title and author list, can be sent to: