Further work

  • Better understanding of energy budget
    • Likely to require some additional global sums
  • Long runs of hybrid model
    • Catch any fields which continue to drift, maybe
      • snow
      • vegetation types, e.g. more broadleaf in one than other
    • These should be fixable, but we could re-start Jnr from Snr's UM dump on say a yearly timescale.
  • I think we'll want to do vertical averaging onto ETA levels for two reasons
    1. The ETA levels on Jnr will not exactly match those on Snr
    2. Eventually we'll want to reduce the vertical levels in Jnr
      • But not near the surface (this is important for emissions)