Skip to content

Generating MOM6 Initial Conditions Using WOA23 dataset

This guide outlines the steps to generate initial-condition fields for MOM6 from World Ocean Atlas 2023 (WOA23) data. The workflow produces 3D fields of:

  • Conservative temperature
  • Potential temperature
  • Absolute Salinity
  • Practical Salinity

on a provided MOM6 grid for use as initial conditions and for model evaluation.

Note

MOM6 always expects initial conditions of potential temperature and practical salinity - see here.

Repository and Requirements

Clone the repository that includes all necessary tools and submodules:

git clone --recursive https://github.com/ACCESS-NRI/initial_conditions_access-om3.git
cd initial_conditions_access-om3

A recursive clone is needed because this repository includes Nic Hannah’s ocean-ic code as a submodule, used to interpolate WOA23 data onto MOM6 3D grids.

Step 1: (Optional) Regenerate Temperature & Salinity from Raw WOA23

Use this step only if you want to regenerate T/S fields from a different version of World Ocean Atlas dataset.

./inte.csh

The inte.csh script processes World Ocean Atlas 2023 (WOA23) data to create consistent monthly temperature and salinity fields suitable for generating MOM6 initial conditions. The WOA23 data used by this script is located at /g/data/av17/access-nri/OM3/woa23.

Purpose

By default, WOA23 provides: - Full-depth annual mean data (XX = 00) - Monthly data (XX = 01–12) only for the upper 1500 m - Seasonal data (XX = 13–16) covering full depth

To use full-depth data with monthly resolution, inte.csh reconstructs it by combining seasonal deep data with monthly shallow data. The WOA23 dataset is based on oceanographic observations collected over the period 1955 to 2022.

What the Script Does

  1. Extract salinity (s_an) from seasonal full-depth data and expand it to monthly resolution.
  2. Use ncks --mk_rec time to add an unlimited time dimension to make the NetCDF files record-aware.
  3. Rename the salinity variable from s_an to practical_salinity for compatibility with the processing pipeline.
  4. Run setup_WOA_initial_conditions.py to merge the monthly upper-ocean data with the seasonal lower-ocean data. During this step, the GSW toolbox is used to calculate:
  5. absolute salinity from practical salinity and pressure
  6. potential and conservative temperature from in-situ temperature, absolute salinity and pressure

The processed monthly files are output to: 

/g/data/ik11/inputs/access-om3/woa23/monthly/<YYYY.MM.DD>

Note

You only need to run this script if you're updating or modifying the WOA23 dataset or prognostic form of temperature or salinity. Otherwise, skip this step and proceed directly to regridding using make_initial_conditions.sh.

Step 2: Regrid to MOM6 Grid

Use the regridding script to interpolate temperature and salinity to your MOM6 model grid.

cd initial_conditions_WOA/
qsub -v VGRID="<path_to_vgrid_file>",HGRID="<path_to_hgrid_file>",INPUT_DIR="<path_to_input_directory>",OUTPUT_DIR="<path_to_output_directory>" -P $PROJECT make_initial_conditions.sh

Replace the variables with your grid and directory paths:

  • VGRID: MOM6 vertical grid file
  • HGRID: MOM6 horizontal grid file
  • INPUT_DIR: Directory with processed WOA23 monthly files
  • OUTPUT_DIR: Where regridded output will be saved

Warning

This code will be very slow for high resolution grids if the regridder's apply_weights function is not built with Pythran - see here

Step 3: Finalize & Tag Metadata

Once satisfied with the output, run:

./finalise.sh -o /g/data/ik11/inputs/access-om3/woa23/025/<YYYY.MM.DD>

This:

  • Commits any changes with git
  • Adds Git metadata to the NetCDF metadata

Notes

  • Make sure to adjust paths in inte.csh if you change the WOA23 dataset.
  • You can repeat Step 2 to generate initial conditions for any resolution. Step 1 is agnostic of the model resolution as the data is on the WOA23 grid.