Create variable mapping dictionaries for GFDL/NCAR/CAM

[durack1]

It seems with the suite of demo scripts that we're starting to collate, it would be suitable to further extend the metrics package with known variable mappings - as has already been done with GFDL and NCAR/ACME/CAM.

It we generate these dictionaries and drop them somewhere (maybe in the doc main subdir, maybe further down under src/python/pcmdi/scripts/) then they can easily be reused/refined as model versions are tweaked - I would be surprised if once we get to the UKMetOffice/ACCESS folks a generic "UM" mapping file can also be created.

An example of this could be:

Model identifier	CMIP variable	Realm	Model variable name	Model units	Conversion algorithm
NCAR-CAM5	pr	atmos	PRECC, PRECL	m s-1	PRECC + PRECL and unit conversion
NCAR-CAM5	rlut	atmos	FSNTOA, FSNT, FLNT	W m-2	FSNTOA-FSNT+FLNT
GFDL-CM4	sos	ocean	SALT	1e-3	1
GFDL-CM4	tos	ocean	TEMP	degrees_C	1
GFDL-CM4	zos	ocean	SSH	m	1

Some of the example lists are linked below: https://github.com/PCMDI/pcmdi_metrics/blob/master/src/python/pcmdi/scripts/GFDL/createXmlAndNcs.py#L58-59 https://github.com/PCMDI/pcmdi_metrics/blob/master/src/python/pcmdi/scripts/ACME/createXmlAndNcs.py#L47-56 https://github.com/PCMDI/pcmdi_metrics/blob/master/src/python/pcmdi/scripts/NCAR/createXmlAndNcs.py#L45-56

[durack1]

@gleckler1 it would be a good idea to go through this data "conditioning" process with @jservonnat and include an IPSL mapping dictionary in our demo templates

[jservonnat]

I'm ok with that too!

[chengzhuzhang]

@durack1 This is exactly what we are looking to do! Should be very helpful for runs without being CMORIZED. I will keep you in the updated as this progresses. Thanks!

[durack1]

@chengzhuzhang ok let's reopen this one.. I think it's a worthy effort

[durack1]

@jservonnat wanna pipe in with your IPSL mappings - have you done this?

[gleckler1]

Good idea to put them in a dictionary json file

[lee1043]

@durack1 Just for housekeeping, I wonder if this issue addressed and okay to close.

[durack1]

@lee1043 good question, to a point this is included in the https://github.com/PCMDI/pcmdi_metrics/tree/main/sample_setups/external-setups examples, but is not managed in any coherent way across these examples (which was the initial intent). As this originates from 9 years ago, would be fine to close

[lee1043]

@durack1 thanks for your comment. I am closing it now but we shall reopen in the future if needed.