Date: 21 August 2024
Attendees:
- Andrew Kiss (AK) @aekiss
- Anton Steketee (AS) @anton
- Ezhil Kannadasan (EK) @ezhilsabareesh8
- Dougie Squire (DS) @dougiesquire
- Siobhan O’Farrell (S0) @sofarrell
- Minghang Li (ML) @minghangli
Minghang:
- Began running tests of 025deg ryf with ACCESS-OM3. 5 year runs, and ML showed plots of Global Mean Ocean T with different parameters
- Tested:
- DTBT_RESET_PERIOD
- DIABATIC_FIRST
- DT_THERM
DT_THERM did show some differences, with larger DT_THERM showing less drift. Its hard to assess as the model hasn’t equalised yet. AK suggested looked at other diagnostics (e.g. those which assess spatial variance), and maybe trying an even bigger DT_THERM. ML will check the tracer diffusion terms in his results.
Dougie:
Updating model components.
- Should we update model components, or move to access-nri forks
- Should we move to the access-nri build system first ?
We will probably move to having forks directly from the main forks (CICE-consortium and MOM-ocean). Well stick with the COSIMA spack for this build.
Next step is to capture on github the steps needed to update our ACCESS-NRI forks. AS will talk to Kieran to change the CICE fork & DS will check with Angus re the MOM fork. See Update components for v0.4.0 · Issue #209 · COSIMA/access-om3 · GitHub
Ezhil:
Has been looking at jumps / gaps / discontinuities in WOA13 and WOA23. They seem to be better in WOA23 but still exist in some places. There is a smoothing step access-om2 generation of initial conditions, we need to investigate further if we still need this step. See Use WOA2023 initial condition · Issue #161 · COSIMA/access-om3 · GitHub
AK also suggest doing some comparisons runs on the equation of state. See MOM6 equation of state · Issue #20 · COSIMA/access-om3 · GitHub