A repository with Jupyter notebooks and datasets that reproduce the analyses and figures included in the paper:
Constantinou, N. C. and Hogg, A. McC. (2021) Intrinsic oceanic decadal variability of upper-ocean heat content. J. Climate. 34 (25), 6175-6189. doi:10.1175/JCLI-D-20-0962.1
@article{Constantinou-Hogg-2021,
doi = {10.1175/JCLI-D-20-0962.1},
year = {2021},
volume = {34},
number = {15},
pages = {6175-6189},
author = {Navid C. Constantinou and Andrew McC. Hogg},
title = {Intrinsic oceanic decadal variability of upper-ocean heat content},
journal = {J. Climate}
}-
prepare-raw-data.ipynb: Downloads raw data from various ACCESS-OM2 models and from CMEMS, manipulates the data (regrids, coarsens, monthly averages, 2D reductions), and produces smaller datasets saved inoutputdirectory. -
grids.zip: compressed.ncfiles with grid details for the ACCESS-OM2 models at three resolutions. Unzip before running notebooks. -
interpolation_weights.zip: compressed.ncfiles with the interpolation weights thatxesmfpackage uses to regrid the ACCESS-OM2 model output and the CMEMS satellite altimetry onto a regular longitude-latitude grid with 1 degree lateral resolution. Unzip before running notebooks. -
output: Empty directory that is populated with output produced byprepare-raw-data.ipynb. Alternatively, if you don't have access to the raw the ACCESS-OM2 model output, you can dowload the processed output from the Zenodo repository at doi:10.5281/zenodo.4924968 and populate theoutputdirectory. -
process-data.ipynb: Process output fromoutputdirectory to produce figures. -
figures: Figures produced byprocess-data.ipynbnotebook.
Some of the raw ACCESS-OM2 model output is available at doi:10.4225/41/5a2dc8543105a. However, accessing output from this doi requires users to have access the Australian National Computing Infrastructure (NCI).
Users with access to NCI's HPC systems should be able to download all model output used
here given that they are members of projects cj50 and ik11. The experiments we use here
are summarized in the table below.
| Resolution | Forcing | Experiment name |
|---|---|---|
| 1 | IAF | 1deg_jra55v13_iaf_spinup1_B1 |
| 1 | RYF | 1deg_jra55_ryf9091_gadi |
| 0.25 | IAF | 025deg_jra55v13_iaf_gmredi6 |
| 0.25 | RYF | 025deg_jra55_ryf9091_gadi |
| 0.10 | IAF | 01deg_jra55v140_iaf, 01deg_jra55v140_iaf_cycle2, 01deg_jra55v140_iaf_cycle3 |
| 0.10 | RYF | 01deg_jra55v13_ryf9091 |
(The IAF 0.10 degree outputs are also available via THREDDS at https://dapds00.nci.org.au/thredds/catalogs/cj50/access-om2/raw-output/access-om2-01/access-om2-01.html.)
Notebook prepare-raw-data.ipynb utilizes cosima-cookbook's
functionality to load the required variables needed from the above-mentioned experiments, do some processing
(e.g., slice and regridding) and then save them as .nc files in the output directory.
If getting access to the Australian National Computing Infrastructure
seems cumbersome, then you can download the processed output from the Zenodo repository
doi:10.5281/zenodo.4924968 and use this sliced/regridded data to
populate the output directory and proceed with analysing it.
After you have unzipped all files and either have run prepare-raw-data.ipynb notebook or
downloaded output from Zenodo repository at doi:10.5281/zenodo.4924968,
the repository's directory structure should look like:
.
├── README.md
├── LICENSE
├── .gitinore
├── grids
│ ├── ocean_grid_10.nc
│ ├── ocean_grid_025.nc
│ └── ocean_grid_01.nc
├── interpolation_weights
│ ├── bilinear_tracer_weights_in025degcmems_out1deg.nc
│ ├── bilinear_tracer_weights_in1degACCESSOM2_out1deg.nc
│ ├── bilinear_tracer_weights_in025degACCESSOM2_out1deg.nc
│ └── bilinear_tracer_weights_in010degACCESSOM2_out1deg.nc
├── output
│ ├── README.md
│ ├── cmems-monthlymean-regridded
│ │ ├── ssh-XXXX.nc
│ │ └── ...
│ ├── ssh-1deg-RYF-regridded
│ │ ├── ssh-XXXX.nc
│ │ └── ...
│ ├── ...
│ ├── ssh-1deg-IAF-regridded
│ │ ├── ssh-XXXX.nc
│ │ └── ...
│ ├── ...
│ ├── uohc-1deg-0m-50m-RYF
│ │ ├── uohc-XXXX.nc
│ │ └── ...
│ ├── ...
│ └── uohc-1deg-0m-50m-IAF
│ ├── uohc-XXXX.nc
│ └── ...
├── figures
│ ├── figure1.pdf
│ └── ...
├── prepare-raw-data.ipynb
└── process-data.ipynb