Daniel Huppmann

I think https://github.com/IAMconsortium/pyam/pull/637 would be helpful for implementation... To avoid too much filtering. Hope I get to review this later today...

I think some configuration like the following would be useful in a `nomenclature.yaml`: ```yaml - time: year: true datetime: true ``` or ```yaml - time: year: true datetime: UTC+01:00 ```...

Note that validation of "subannual" (categorical) timesteps already works simply by adding a folder `subannual` and activating it via `nomenclature.yaml`, see [here](https://github.com/openENTRANCE/openentrance/tree/main/definitions/subannual) in the openentrance project.

Correct, this would allow all three pyam-supported time domains.

Well… There are three types of time format in pyam and ixmp4 called “year”, “subannual” and “datetime”, so I would keep the naming in nomenclature consistent. But if you think...

As just discussed, let's first only validate the timezone. So if ``` time: datetime: UTC±HH:MM ``` and the IamDataFrame has a datetime or mixed time domain, then the datetime items...

Good observation, but I don't think this is a high-priority corner case... I'd keep that on a list for "implement when the first use case emerges".

And a quickfix is already included as part of https://github.com/IAMconsortium/pyam/pull/868 - question is whether we need this as a pyam release?

Seems that this issues was fixed just now: https://github.com/IAMconsortium/pyam/pull/869

I believe that you refer to this warning: > No model mapping found for model ... I'd still keep it but rephrase to > Skipping region aggregation for model '{}'...