How can we include `NorESM1`
has transient CO2, areosol and volcanic forcing. all other models are (esm-)piControl.
- detrending, but how properly?
I always get spurious trends and late predictable leads.
Discussed this at length with Filippa at EGU22 in Vienna. Fransner et al. use a different approach to calc prediction skill, therefore results not comparable: https://climpred.readthedocs.io/en/stable/examples/decadal/verify_dim_implications.html
2 datasets:
- initialized
- historical: only on levante so far