CMIP7 custom table entry

[hb326]

For the REF we have defined several ozone diagnostics that are based on a 3D ozone field. This field was requested by the REF data request. For most parts of the REF ozone recipe the variable toz is calculated from the 3D field. But this variable is not defined in the same CMOR table as the 3D ozone field: toz is defined in aerosol whereas o3 is defined in atmosChem.

When I want to use the defined toz variable, the branded suffix for the data file will be wrong (compared to the branded suffix of o3). But if I want to calculate toz from o3, I don't have a definition for the variable in the CMOR table.

Do I "just" start the custom CMOR table with this variable?

[schlunma]

It might work to adapt the get_required method of the derivation class so that it automatically changes the relevant facets (branding_suffix and mip):

ESMValCore/esmvalcore/preprocessor/_derive/toz.py

Lines 66 to 81 in edb6b2c

	@staticmethod
	def required(project):
	"""Declare the variables needed for derivation."""
	# TODO: make get_required _derive/__init__.py use variables as argument
	# and make this dependent on mip
	if project == "CMIP6":
	required = [
	{"short_name": "o3"},
	{"short_name": "ps", "mip": "Amon"},
	]
	else:
	required = [
	{"short_name": "tro3"},
	{"short_name": "ps"},
	]
	return required

[bouweandela]

As @schlunma suggests, this looks like a case of #2984. However, I don't quite understand why you want to calculate toz from o3 if you can just download files which already contain toz. Was there a mistake in requesting the input variables needed for the REF?