In order to use PrePARE please follow these instructions.

PrePARE has been created to validate CMIP6 data before publishing files to ESGF. It may not work properly on CMIP5 files.


PrePARE  [-h] [-l [CWD]] [--variable VARIABLE] [-v] [--table-path TABLE_PATH] 
         [--max-processes 4] [--all] [--hide-progress] [--no-text-color] 
         [--ignore-dir PYTHON_REGEX] [--include-file PYTHON_REGEX] 
         [--exclude-file PYTHON_REGEX]
         input [input ...]


  • input Input CMIP6 netCDF data to validate. If a directory is submitted all netCDF recursively found will be validated independently.

  • -h Display synopsis of the program.

  • -l, –log Logfile directory. Default is the working directory. If not, standard output is used. Only available in multiprocessing mode.

  • –variable Specify geophysical variable name. If not variable is deduced from filename.

  • -v, –version Version of software.

  • –table-path Specify the CMIP6 CMOR tables path (JSON file). If not submitted read the CMIP6_CMOR_TABLES environment variable if it exists. If a directory is submitted table is deduced from filename (default is “./Tables”).
  • –max-processes Maximum number of processes to simultaneously validate several files. Set to one seems sequential processing (default). Set to “-1” uses all available resources as returned by “multiprocessing.cpu_count()”.

  • –all Show all results. Default only shows error(s) (i.e., file(s) not compliant).

  • –hide-progress Do not show the percentage of progress / number of files checked while running PrePARE.

  • –no-text-color Remove text color from output.

  • –ignore-dir Filter directories NON-matching the regular expression. Default ignores paths with folder name(s) starting with “.”.

  • –exclude-file Filter files NON-matching the regular expression. Duplicate the flag to set several filters. Default only exclude hidden files (with names not starting with “.”).


PrePARE will verify that all attributes in the input file are present and conform to CMIP6 for publication into ESGF. We also recommand running the python program cfchecker created by the University of Reading in the UK to confirm that your file is CF-1 compliant.

  • In order to validate all CMIP6 required attributes by PrePARE, a Controlled Vocabulary file is read by the program where a JSON dictionnary called “required_global_attributes” point to a list of strings. Each element of that list corresponds to a global attribute.
  • PrePARE can also use regular expressions to validate the value of the some global attributes. Here is an example used for variant_label.

  • Institutions and institution_ids need to be registered into a list. PrePARE will only accept institutions which have been pre-registered for CMIP6 publication into ESGF. Click here for the list of institutions. If you wish to register your institution write to the cmor mailing list.

  • Source and Source ID also need to be registered for CMIP6 publication. Here is the list of registered sources.

  • Only experiments found in the Controlled Vocabulary files are accepted for CMIP6 publication. A list of experiment_ids have been pre-defined including mandatory attributes. A warning will be displayed if one experiment attribute is missing or is not properly set by your program.

  • grid and nominal_resolution are mandatory global attributes in CMIP6. PrePARE will make sure that these attributes are conformed to one of the following syntax:

    grid nominal_resolution
    gs1x1 1x1 degree
    gs1x1 or gn0 to gn9 1x1 degree
    gs1x1 or gr0 to gr9 1x1 degree
    gn0 to gn9 “5 km” or “10 km” or “25 km” or “50 km” or “100 km” or “250 km” or “500 km” or “1000 km” or “2500 km” or “5000 km” or “10000 km”
    gr0 to gr9 “5 km” or “10 km” or “25 km” or “50 km” or “100 km” or “250 km” or “500 km” or “1000 km” or “2500 km” or “5000 km” or “10000 km”
  • PrePARE verifies that the creation date found in the netCDF file is conform to ISO 8601 standard.

  • The Further Info URL attribute has to be set according to a very specific template. The PrePARE will use global attribute names found in the netCDF input file and replace the corresponding tag found in a template to rebuild proper CMIP6 link. If the reconstructed URL does not correspond to the value found in the input file, PrePARE will display an critical error on the screen.
    • Below is the defaul template used for the furtherinfourl attribute. Each string found between the “<>” characters correspond to a global attribute. The program will replace these strings with the corresponding global attribute values and add the ”.” character between each tag.<mip_era><institution_id><source_id><experiment_id><sub_experiment_id><variant_label>

  • PrePARE will also verify variable attributes necessary for CMIP6 publication. It validates: long_name, standard_name, units and missing_value according the CMIP6 tables information.
Tags: cmip6