Add a New MIP tables Cookbook?

[jypeter]

You may also want to add PCMDI/cmor#255 and similar information to a Creating new tables (for a new MIP?) cookbook in the Appendix

[durack1]

The demo at PCMDI/obs4MIPs-cmor-tables/demo and the example of the input4MIPs tables https://github.com/PCMDI/input4MIPs-cmor-tables should be good things to link to - they're examples in place, so users can use these immediately

[jypeter]

So that's the kind of script I would have to use in order to prepare PMIP boundary conditions data for input4MIPs ?

[durack1]

@jypeter if you wanted to create PMIP data for input4MIPs, we could augment the existing input4MIPs-cmor-tables to account for that, so these are maintained into the future for reuse.

[wachsylon]

We need a manual about how to create CMOR-MIP-tables. Here are some thoughts about it:

1. Project's MIP-tables should be at least linked. This helps data request as well as project's manager to gain an overview about what similar projects plan to produce. It should be possible to identify variable duplicates from other projects.
   One way to do that is to collect Project's MIP-tables in one directory. The history of the table creation might also be well retraceable when the new table's repository is forked from the original CMIP6-MIP-table repository.
2. The naming convention should be explained. An idea for the future is to have all prefix, suffix and qualifier within a mip-table name everytime (3hrMean and 3hrPt instead of 3hr and 3hrPt).
3. Grids and coordinates tables probably should not be changed. If someone needs new coordinates, they probably should be included in the original CMIP-MIP-tables.
4. If there is a software/script to write MIP-tables based on a DReq file, it would be very helpful to see it.
   Best regards,
   Fabi