Porting essential NCL diagnostics

[axel-lauer]

Since NCL is not being maintained any more, it might eventually become unfeasible to include NCL in the ESMValTool environment (see also #1057). This will make it probably rather difficult if not impossible to run NCL diagnostics with future releases of the ESMValTool. Some NCL diagnostics, however, are deemed important enough to make the effort to rewrite these diagnostics e.g. in Python or another supported language such as R.

The aim of this discussion is to create a list of such NCL diagnostics and assign a priority (1: essential, 2: important, 3: nice to have) based e.g. on scientific value or outstanding project deliverables.

List of NCL diagnostics and CMORizers to be converted

recipe / diagnostic / cmorizer	priority
recipe_perfmetrics	1
recipe_ipccwg1ar6ch3_atmosphere.yml	2
recipe_ipccwg1ar6ch3_fig_3_19.yml	2
recipe_ipccwg1ar6ch3_fig_3_43.yml	2

[bettina-gier]

recipe_smpi also uses the perfmetrics diagnostics, do we keep these together? Implemented with a lot of if clauses, so maybe on porting see what should be generalized and put into separate files

[LisaBock]

Finally I also collected a list of IPCC diagnostics which, I think, are needed to be available in Python:

• bias map plots with stippling/hatching of the model agreement regarding the consideration of internal variability (IPCC AR6 WG1 Fig. 3.3, 3.13),
• anomaly time series of near-surface air temperature with volcano eruptions and absolute mean temperature attached (IPCC AR6 WG1 Fig. 3.4),
• precipitation anomaly - timeseries and trend zonal ((PCC AR6 WG1 Fig. 3.15),
• mean wind - absolute and linear trend (IPCC AR6 WG1 Fig. 3.19),
• pattern correlation figure (IPCC AR6 WG1 Fig. 3.43)

The basis of these plots should be no problem but they all need some extras and/or specific layout requirements.

@axel-lauer - Shall I add them to the table above?

[axel-lauer]

Yes, if possible, please add the diagnostics to the table above.