ESMValTool v2.12.0 released!
New versions of the ESMValTool (v2.12.0) and ESMValCore (v2.12.0) have just been released:
- https://github.com/ESMValGroup/ESMValTool/releases/tag/v2.12.0
- https://github.com/ESMValGroup/ESMValCore/releases/tag/v2.12.0
These versions are available, as usual, on GitHub, conda-forge, PyPI and as containers.
Please refer to the Changelogs for an overview of the latest changes and additions:
- https://docs.esmvaltool.org/en/latest/changelog.html#v2-12-0
- https://docs.esmvaltool.org/projects/esmvalcore/en/latest/changelog.html#v2-12-0
Highlights:
ESMValCore:
- Preprocessor esmvalcore.preprocessor.extract_time() now allows to extract time blocks in each year by making parameters start_year and end_year optional.
- A new way of configuring the tool has been developed.
- Performance improvements:
- An iris-esmf-regrid scheme has been added to preprocessor esmvalcore.preprocessor.regrid(), which improves the regridding of 2D grids and adds the capability to regrid UGRID meshes out of the box.
- Data is now saved from one preprocessing task at the time when using the distributed scheduler, in order to avoid running out of memory.
- A better default num_workers has been set when using more than one max_parallel_tasks with an unconfigured threaded scheduler.
- An on-the-fly cmorizer for ACCESS native data is now available.
ESMValTool:
- Recipe recipe_portrait_CMIP.yml allows the creation of portrait plots to visualize performance metrics.
- A set of new recipes allow plotting arbitrary preprocessor output to benchmark a model simulation with other datasets. See also ESMValTool paper on benchmarking and monitoring climate models.
Happy model evaluation with ESMValTool!