ACCESS-coupled N48 issue #1: change topography, bathymetry and land-mask

Dear ACCESS-NRI team,
How can we change topography, bathymetry and land-mask in the ACCESS-coupled N48?

As a starting point I would suggest we just change a dummy test: move Australia to another location in the Pacific.

Please, tell us how is this down?

best regards
Dietmar

@atteggiani @dkhutch

1 Like

Hi @Dietmar_Dommenget,

To create the land-mask files (for both ocean and atmosphere grid) from the bathymetry file, you will need to use a tool (make_topog) available inside the ACCESS-OM2 repo.

I would suggest first step is to produce those files to run a Ocean-only (MOM standalone) or maybe Ocean-SeaIce (ACCESS-OM2) simulation with your changed continent displacement.
Then, after that step is done, you can produce the atmosphere ancillary files (based on the land-mask file you created) and run a coupled simulation.

I am going to do this exact step so I can learn more about the ocean input files and how the whole process works.
I will start by running an ACCESS-OM2 (1° with 50 vertical levels) historical simulation using pre-made input files, and then repeat the same kind of simulation by creating my own (ocean) input files from scratch (starting with the creation of land-mask file from bathymetry). I will keep you updated on the entire workflow.

Cheers
Davide

3 Likes

Thanks Davide! Further to Davide’s comments. The ‘make_topog’ tool is found within the MOM5.1 repository (part of the ACCESS OM2). You can see the tool here:

‘make_topog’ takes an input file of ocean bathymetry (0 is land, depth is positive downwards) and interpolates it onto the ocean grid (defined by ocean_hgrid.nc)

I have used this on Gadi to do novel topographies using the GFDL CM2.1 coupled model (using the same ocean model).

I don’t know how to do the ancillary files for the atmosphere, or the land model boundary conditions, hence those are the parts I’m hoping to learn.
@Dietmar_Dommenget

2 Likes

Note that this is also being discussed in this topic. Not exactly the same problem, but there is significant overlap.

1 Like