ZEC diagnostic: consider supporting esm-1pctCO2 as a reference experiment

[lewisjared]

Context

The ZEC (Zero Emission Commitment) diagnostic requires a reference experiment alongside the esm-1pct-brch-1000PgC child experiment. Currently, the ESMValTool climate_metrics/zec.py script only accepts 1pctCO2 and esm-flat10 as valid reference experiments.

Problem

Some CMIP6 models branch esm-1pct-brch-1000PgC from esm-1pctCO2 rather than 1pctCO2:

Model	parent_experiment_id
GISS-E2-1-G-CC	esm-1pctCO2
MIROC-ES2L	esm-1pctCO2
MPI-ESM1-2-LR	esm-1pctCO2

The ESMValTool ZEC diagnostic rejects esm-1pctCO2 with:

ValueError: esm-1pctCO2 is not a valid experiment for calculating ZEC
Current accepted experiments are {'reference': ['esm-flat10', '1pctCO2'], 'simulation': [...]}

Current workaround

In #585, we added a RequireFacets("experiment_id", ("1pctCO2",)) constraint to the CMIP6 ZEC data requirement, which filters out models whose parent is esm-1pctCO2 during constraint solving. This prevents failed executions but reduces model coverage.

Questions for upstream

1. Is esm-1pctCO2 scientifically valid as a reference experiment for ZEC? If so, ESMValTool's climate_metrics/zec.py should add it to the accepted reference experiments.
2. If not, should these models be excluded from ZEC analysis entirely, or is there an alternative approach (e.g., using 1pctCO2 data from the same model even though it's not the metadata-declared parent)?

Affected models

These models have esm-1pctCO2 as parent for CMIP6, but note that the CMIP7 data path (which uses RequireFacets on both experiments directly) still includes them because CMIP7 resolves 1pctCO2 separately.

[bouweandela]

The experiments are configurable, I've made a pull request to update the way we write the recipe.