Jupyter notebooks to reproduce figures for the paper:
Bhagtani, D., Hogg, A., McC., Holmes, R. M., and Constantinou, N. C. (2023) Surface heating steers planetary-scale ocean circulation. J. Phys. Oceanogr., 53(10), 2375–2391. doi:10.1175/JPO-D-23-0016.1
Figures: Figures produced by jupyter notebooks in home directory.
NETCDF: Data files containing:
Gyre_95th: gyre strength estimated using isopycnal outcropping method, andSST_expts: pre-calculated globally integrated sea surface temperatures used for Fig. 1.
Expt_config_figures.ipynb: Notebook to create Fig. 1.
Wind_figures.ipynb: Notebook to create Figs. 2 and 3.
Red_heat_figures.ipynb: Notebook to create Figs. 4, 5, 6, and 7.
Inc_heat_figures.ipynb: Notebook to create Figs. 8, 9, 10, 11, and 12.
Uniform_warming_figures.ipynb: Notebook to create Fig. 13.
We run ACCESS-OM2 (Kiss et al 2020) at 0.25 degree resolution which comprises of:
(i) MOM v5.1: https://github.com/dhruvbhagtani/MOM5/tree/f3f05e7700ca29ed1baa5f1c8d1f411ae9a40ae3 with LIBACCESS-OM2 modules used: https://github.com/COSIMA/libaccessom2/tree/392dce398e985843cc6488af4c3498138373838b.
(ii) CICE v5.2: https://github.com/COSIMA/cice5/tree/26e61591c985ee76a10366dcba51fb99e6d84f61 with LIBACCESS-OM2 modules used: https://github.com/COSIMA/libaccessom2/tree/a9d4b678166642f0a7639a8a878afcfec1f35c4f.
Please note that the MOM5 code used above has the following additions:
(i) A modified boundary layer scheme (K-profile parameterisation; Large et al 1994) for mixing layer depth, and
(ii) A script to input a mask and place it on top of wind forcing.
This eddy-permitting model is run for 200 years (more details present in Section 2 of the paper), after which we initialise a MOM5-only control simulation for 100 years. This simulation also uses the modified KPP scheme similar to that in the ACCESS-OM2-025 control simulation. After 100 years, we branch off the following set of MOM5-only perturbation simulations (the MOM5-only control simulation is also run for another 100 years):
| Expt short name | Expt long name | Wind Factor | Surface buoyancy flux contrast (W m⁻²) | Region |
|---|---|---|---|---|
| Control | 025deg_jra55_ryf_control |
1 | 0 | G |
| 0.5xW | 025deg_jra55_ryf_fluxS_050x_20yr_avg |
0.5 | 0 | G |
| 1.5xW | 025deg_jra55_ryf_fluxS_150x_20yr_avg |
1.5 | 0 | G |
| -15 W m⁻² | 025deg_jra55_ryf_fluxH_neg10W |
1 | -15 | G - T |
| -7.5 W m⁻² | 025deg_jra55_ryf_fluxH_neg5W |
1 | -7.5 | G - T |
| +7.5 W m⁻² | 025deg_jra55_ryf_fluxH_pos5W |
1 | +7.5 | G - T |
| +15 W m⁻² | 025deg_jra55_ryf_fluxH_pos10W |
1 | +15 | G - T |
| +30 W m⁻² | 025deg_jra55_ryf_fluxH_pos20W |
1 | +30 | G - T |
| Uniform warming | 025deg_jra55_ryf_fluxH_pos5W_globe |
1 | 0, instead spatially uniform +5 | G |
Each flux-forced perturbation experiment is run for 100 years. Outputs for each experiment can be found on the National Computational Infrastructure in the following directory: /g/data/hh5/tmp/db6174/mom/archive. Access to these outputs requires one to be a member of this infrastructure. Therefore, a copy of pre-processed files are available at the following Zenodo repository to reproduce figures: 10.5281/zenodo.8405008.
Finally, the MOM5 version used for the perturbation experiments is: https://github.com/dhruvbhagtani/MOM5/tree/d7d72278a11ed9e2d88be3cb8d780b8efba629c5.
A climatology of the following data is constructed from the last 20 years of the 200 year ACCESS-OM2 simulation to create surface boundary fluxes:
(i) Wind stress (x- and y-direction),
(ii) River runoff and precipitation,
(iii) Surface heat flux components: shortwave and longwave radiation, along with sensible and latent heat fluxes, and
(iv) Time-varying surface salt restoration.
NOTE: To prevent model instability, the surface heat flux components do not contain heat input due to frazil formation, and is dynamically calculated in all MOM5 simulations.
For wind stress sensitivity experiments, we multiply the x- and y-directed wind stress globally by the required factor. For surface buoyancy flux contrast simulations, we apply an anomalous heat flux (shown in Fig. 1b in the paper) on top of the control simulation. The motivation behing the shape and amplitude of the heat flux function is outlined in section 2 of the paper.
- Kiss, A. E., Hogg, A. McC., Hannah, N., Boeira Dias, F., Brassington, G. B., Chamberlain, M. A., Chapman, C., Dobrohotoff, P., Domingues, C. M., Duran, E. R., England, M. H., Fiedler, R., Griffies, S. M., Heerdegen, A., Heil, P., Holmes, R. M., Klocker, A., Marsland, S. J., Morrison, A. K., Munroe, J., Nikurashin, M., Oke, P. R., Pilo, G. S., Richet, O., Savita, A., Spence, P., Stewart, K. D., Ward, M. L., Wu, F., and Zhang, X.: ACCESS-OM2 v1.0: a global ocean–sea ice model at three resolutions, Geosci. Model Dev., 13, 401–442, doi:10.5194/gmd-13-401-2020, 2020.