Dataset for paper Pavel Perezhogin, Laure Zanna, Carlos Fernandez-Granda "Generative data-driven approaches for stochastic subgrid parameterizations in an idealized ocean model" submitted to JAMES.

<p>The dataset consists of the directory tree of .zarr archives. See <a href="https://github.com/m2lines/pyqg_generative/blob/master/Google-Colab/dataset.ipynb">Github repository</a>&nbsp;for the description of the dataset.</p> <p>The directory tree is:</p> <pre><code>├── eddy │ ├── 48 │ │ ├── gauss │ │ ├── hires-gauss │ │ ├── hires-sharp │ │ ├── lores │ │ └── sharp │ ├── 64 │ │ ├── gauss │ │ ├── hires-gauss │ │ ├── hires-sharp │ │ ├── lores │ │ └── sharp │ ├── 96 │ │ ├── gauss │ │ ├── hires-gauss │ │ ├── hires-sharp │ │ ├── lores │ │ └── sharp │ └── hires ├── jet │ ├── 48 │ │ ├── gauss │ │ ├── hires-gauss │ │ ├── hires-sharp │ │ ├── lores │ │ └── sharp │ ├── 64 │ │ ├── gauss │ │ ├── hires-gauss │ │ ├── hires-sharp │ │ ├── lores │ │ └── sharp │ ├── 96 │ │ ├── gauss │ │ ├── hires-gauss │ │ ├── hires-sharp │ │ ├── lores │ │ └── sharp │ └── hires</code></pre> <ul> <li>Every individual dataset is a&nbsp;<code>.zarr</code>&nbsp;<a href="https://zarr.readthedocs.io/en/stable/">archive</a></li> <li><code>eddy/jet</code>&nbsp;- configuration of the pyqg; eddy is default; See&nbsp;<a href="https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2022MS003258">Ross2022</a>&nbsp;for description</li> <li><code>hires.zarr</code>&nbsp;- high-resolution simulation at 256x256 grid</li> <li><code>48/64/96</code>&nbsp;- resolution of the coarse models</li> <li><code>lores.zarr</code>&nbsp;- low-resolution simulation</li> <li><code>gauss.zarr</code>,&nbsp;<code>sharp.zarr</code>&nbsp;- training datasets for prediction of subgrid forcing obtained with Gaussian or Sharp filters</li> <li><code>hires-gauss.zarr</code>,&nbsp;<code>hires-sharp.zarr</code>&nbsp;- high-resolution simulation projected onto coarse grid with Gaussian or Sharp filters</li> </ul> <p>The directory tree is split into small tar.gz files each representing a separate .zarr archive. Download any required parts of the dataset and unpack with:</p> <p><strong>tar -xf *.tar.gz&nbsp;</strong></p> <p><strong>The directory tree will be restored automatically!</strong></p>