Spatial scales
I think we need to consider the spatial scales involved. For UKESM-hybrid
N96 N48 ORCA1
- Lsnr for N96 Snr is ~140km.
- Hence, error in any averaging is ~Lsnr/2 ~ 70km
- Any science >> Lsnr should be OK to average from Snr -> Jnr
- I think U, V, θvd & Π are largely controlled by such
science
- Any science << Lsnr cannot be averaged from Snr -> Jnr
- I think we need to trust Jnr to handle this
- Large-scale cloud/precipitation in Snr is ~Lsnr
- What do we do with science ~Lsnr?
- I'm increasingly believing that we can't average this from Snr -> Jnr
- However, this doesn't mean we give up yet.