Some fixes in the seasonal bias plots

[jesusff]

This is introducing a 0.35K bias, right? the offset for degC to K should be 273.15
https://github.com/euro-cordex/joint-evaluation/blob/58b71dd6b61cd066686ce69731cb0b9165eb9304/eval-book/eobs.ipynb#L1253C29-L1253C34

We should use the source_id as labels, rather than the institution_id (the same institution can run different models)
https://github.com/euro-cordex/joint-evaluation/blob/58b71dd6b61cd066686ce69731cb0b9165eb9304/eval-book/eobs.ipynb#L1259

[larsbuntemeyer]

Agreed! I'll update as soon as possible. Do you have an opinion on #9 ?

[jesusff]

yes, I'd go for the missings threshold to get rid of those awful artifacts in the biases