Features#

auto-icon provides already a large set of features. The most prominent ones are:

General features#

  • Manage multiple simulation runs with simimlar configuration all at once.

  • Manage a variety of ICON(-ART) workflows.

  • Flexibly include pre- and post-processing into the workflow - all tasks are triggered automatically.

  • Reinitialize your model run regularly with new meteorology, aerosol or chemistry data.

Input data, pre- and post-processing#

  • Use existing grid files from DWD and MPIM server by specifying only R, B and grid ID.

  • Remapping of arbitrary data as input to the desired ICON grid, specifically IFS and ERA5 data.

  • (Levante only): automatic retrieval of ERA5 input data for remapping: start date and grid spec (as above) is the only required input.

  • Do (nearly) arbitrary post-processing with CDO, specified by a simple configuration interface.

  • Run sanity checks with the SAMOA tool (access to SAMOA required).

Configuring ICON - Namelist features#

  • Set default values for output_nml’s applying to all sections where not explicitly overwritten (e.g. file_iterval).

  • Full icon_master.namelist is written by auto-icon without extra user input.

  • Automatic substitution of

    • Output start and end dates to simulation start/end.

    • Filetypes for input and output files.

    • Model timing.

Feature outlook#

This is a non-complete and non-ordered list on planned features. More detailed information on the actual development process can be found in the issues section.

  • Native support for multiple sources of input data (e.g. online repositories, data on different HPC systems)

  • Integration of the ESMValTool.

  • Suite of test and benchmark jobs.

  • Connection to DWD’s basic cycling tool BaCy.

Missing a feature?

Please file an issue (DKRZ internal or public).