Migration to NEMO v4.2

[jdha]

Begin migration to NEMO v4.2

• [ ] Update setup script to pull from NEMO gitlab (@jdha - done in other repos)
• [ ] Re-organise repo structure to accommodate different versions of NEMO (@jdha - done in other repos)
• [x] Migrate 4.0.4 namelist_cfg to 4.2 (@anwiseNOCL ?)

Testing

• [x] switch off LBCs, SBCs, ICs run with CO9 domain_cfg and usr_def_istate
• [ ] switch on LBCs (equal to ICs) :heavy_check_mark: 1. with 'old' bdy coordinates files used in previous version
  2. with 'new' bdy coordinates files created by PyNemo
• [ ] add in JRA SBCs and document setup (scripts to assist?)
• [ ] add in realistic (GO8p6) ICs and document setup (scripts to assist?)
• [ ] add in realistic (GO8p6) LBCs and document setup (scripts to assist?)
• [ ] switch on tides
• [ ] add option to setup script (and test) for XIOS3 for NEMO>=4.2.0

Things to bare in mind

We start with an empty MY_SRC. This may have to change to get things running. Once we have a working version a discussion needs be had about what, if anything, additions need porting to 4.2.

[anwiseNOCL]

@jdha @jpolton I've migrated the namelist from 4.0.2 to 4.2.0. There are a number of changes that will need to be tested.

Once we have the repo sorted I can add the namelist, which I have annotated.

[jdha]

@anwiseNOCL I'll endeavour to finish my two tasks by the end of the week!

[anwiseNOCL]

@anwiseNOCL I'll endeavour to finish my two tasks by the end of the week!

@jdha I couldn't resist having a tinker with v4.2 and it's been fairly clean sailing. I have AMM7 V4.2.0 running with no forcing and a horizontally uniform density. I tried it out with a Siddorn Furner analytic profile with constant salinity and depth dependent temperature that I added in V4_0 for the vert. coord stuff.I haven't looked in detail at output but it runs fine so that is good news. I used the gnu compiler. The cray compiler worked but I got really unhelpful error messages when I ran the model with a typo in the namelist so abandoned it.

[jdha]

@anwiseNOCL good to know that it works - just tidying up the setup script to give options of 4.0.2, 4.0.4 and tagged 4.2 (adding a few more options). I'm folding in a few updates to runscripts and envs from SE-NEMO. So hopefully I'll be able to tick my two tasks off by Monday.

[anwiseNOCL]

@jpolton @jdha The move to V4.2 is linked to creating the 1960 to present-day CANARI run. I want to completely document the process of setting up AMM7 for this. The wiki could be a sensible place for this, except that it has all the vertical coordinates stuff in it. How can we structure the repo to support this?

[jdha]

@anwiseNOCL I would just add a top layer in the wiki. e.g.

Home

• Vertical Coordinate Study (put everything in the current wiki under this)
• AMM7 CANARI Development

[jpolton]

@anwiseNOCL I would just add a top layer in the wiki. e.g.

Home

• Vertical Coordinate Study (put everything in the current wiki under this)
• AMM7 CANARI Development

Maybe add some dates to these pages, so that people can see what is current and what is historical