Jobs left to do

Science still to be added

  • Nitrogen limitation (nitrogen scheme became available at UM10.5)
  • Crop and pasture forcing (not in UM10.5)
  • Spencer is building an atmosphere only job, using the above, at UM10.5.

Ancillaries needed

  • From Eddy
    • Crop fraction
    • Pasture fraction
  • From Spencer
    • Nitrogen deposition
  • From Rich (converting ancillaries on a ORCA025 grid to ORCA1)
    • Fraction of surface type
    • Leaf area index
    • Canopy height
    (NOTE: may need to spin-up Jules + coupled run, both with ORCA1 grid, to properly create these files later)

MEDUSA

  • Check conversion of CO2 flux factor
  • NRUN-CRUN fix?
    • mocsy (calculates the surface ocean's carbon system) is only called from ocean interier every month
  • Put input files into central repository (reduce size of UKESM_fields_1860_eORCA1.nc first)
  • Maybe add input files into MOCI branch (as Tim Graham has done for GO6)?

Performance of prototype UKESM

  • Fix/understand why the months that follow 1st month are much slower, e.g. u-af826
    • 23 nodes (20 atmos, 3 ocean)
    • 1st month: 1.7 model years/day
    • Following months: 1.2 model years/day
  • Find why OpenMP does not bit-reproduce non-OpenMP results

Rose-stem job for UKESM

  • Rose-stem jobs cannot have branches
  • Submitting MEDUSA currently requires a branch
  • The submission code is likely to be replaced in the next few months by Harry Shepherd's driver files
  • Job options:
    • a full chemistry coupled job with TRIFFID, or
    • we could be put GC3-coupled file in suite (we're not sure if this is allowed)