• Python 3.*
• Python packages numpy, scipy, netCDF4 and matplotlib
• GMAP tools in /home/common/sync

To install EMS on a CNRM computer, with access to the CNRM Lustre system:

1. Get the installation script:

   wget https://raw.githubusercontent.com/romainroehrig/EMS/master/install.sh

2. Modify install.sh (or execute install.sh -h to get its usage):

   • Set EMS_VERSION, e.g., EMS_VERSION=2.4.5
   • Set where you want to install EMS: default is REP_EMS=$HOME/Tools/EMS/V${EMS_VERSION}
   • Set where you want to run MUSC: default is REP_MUSC=$HOME/MUSC/V${EMS_VERSION}

3. Execute install.sh. A test is done at the end with ARPEGE-Climat 6.3.2 for the ARMCU/REF case.

1. Go in the REP_MUSC directory

2. Source setenv to have the right PATH and PYTHONPATH environment variables

   source setenv

3. You can manage your own namelists (namelist directory), vertical grid (grid directory) and MUSC configuration files (config directory), and postprocessing (post directory).

4. Run MUSC:

   MUSC.py -config config/YOUR_CONFIG_FILE -case CASE -subcase SUBCASE

• ems_list_cases.py: Provide information about cases/subcases available in EMS
• ems_prep_init_forc_atm.py: prepare the nam1D namelist for ARPIFS restart
• ems_prep_nam_atm.py: Modify an ARPIFS namelist to simulate a case described by a netcdf file using the DEPHY common format
• ems_prep_nam_sfx.py: Modify a SURFEX namelist to simulate a case described by a netcdf file using the DEPHY common format
• ems_convert2p.py
• ems_convert2z.py
• ems_convertLFA2nc.py
• ems_lfa2nc.py