v4 preview release #291
There are no files selected for viewing
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,31 @@ | ||
| --- | ||
| title: 'Parcels v4.0-preview is available! 🚀' | ||
| date: '2026-08-01' | ||
|
Member
Author
There was a problem hiding this comment. This is a placeholder for now; will need to adapt this to the actual release date |
||
| authors: | ||
| - name: Erik van Sebille | ||
| github: erikvansebille | ||
| - name: Nick Hodgskin | ||
| github: VeckotheGecko | ||
| - name: Joe Schoonover | ||
| github: fluidnumericsJoe | ||
| summary: 'Parcels v4.0-preview is now available, bringing exciting new features and improvements to the Parcels framework.' | ||
| --- | ||
|
|
||
| We are very excited to announce that Parcels v4-preview is now available for performing Lagrangian simulations in hydrodynamic flows. Parcels v4 is a complete redesign of the [Parcels code](https://parcels-code.org) and is designed to be more flexible and extensible than previous versions. | ||
|
|
||
| Parcels v4 is pure python code, so that it can leverage the full power of the [Pangeo ecosystem](https://pangeo.io/) of geoscience python tools such as xarray, dask and zarr. Particle advection and interpolation are now vectorised operations, which improves performance and makes [particle-particle interactions](https://docs.parcels-code.org/en/main/user_guide/examples/tutorial_interaction.html) easier to implement. | ||
|
erikvansebille marked this conversation as resolved.
Outdated
|
||
|
|
||
| Key improvements of Parcels v4 are | ||
|
erikvansebille marked this conversation as resolved.
Outdated
|
||
|
|
||
| 1. Support for [unstructured grids](https://docs.parcels-code.org/en/main/user_guide/examples/explanation_grids.html). | ||
| 2. More flexibility in writing Kernels (including [using python functions in Kernels](https://docs.parcels-code.org/en/main/user_guide/examples/tutorial_gsw_density.html) .and support for writing output in a Kernel) | ||
| 3. Support for [user-defined interpolators](https://docs.parcels-code.org/en/main/user_guide/examples/explanation_interpolation.html). | ||
| 4. More robust and flexible support for [nested grids](https://docs.parcels-code.org/en/main/user_guide/examples/tutorial_nestedgrids.html). | ||
|
|
||
| We have also changed the [output format to `parquet`](https://docs.parcels-code.org/en/main/getting_started/tutorial_output.html#reading-the-output-file), which is a tabular format and therefore more efficient and flexible for storing large datasets. | ||
|
erikvansebille marked this conversation as resolved.
Outdated
|
||
|
|
||
| See the full migration guide from Parcels v3 to v4 [here](https://docs.parcels-code.org/en/main/user_guide/v4-migration.html). | ||
|
|
||
| Note that Parcels v4-preview is not a stable release yet, and we are still working on improving the performance of the framework and adding new features. We welcome feedback from the community, and encourage users to try out the new version and [report any issues or suggestions](https://github.com/Parcels-Code/parcels/issues). | ||
|
Member
Author
There was a problem hiding this comment. TODO: Add link to new performance tutorial here
erikvansebille marked this conversation as resolved.
Outdated
|
||
|
|
||
| Funded by the [WarmWorld ELPHE project](https://www.geomar.de/en/research/ongoing-projects/project-details/prj/376612?cHash=449a07d057ea0d890287bd2b8ca26e8b) funded by the German Federal Ministry for Research, Technology and Space (BMFTR) and the [NWO Vici project “Tracing Marine Macroplastics by Unraveling the Ocean’s Multiscale Transport Processes”](https://www.nwo.nl/en/projects/vic222025), the development of Parcels v4 started in September 2025 and was a collaboration between the [GEOMAR Helmholtz Centre for Ocean Research Kiel](https://www.geomar.de/en/), [FluidNumerics](https://www.fluidnumerics.com/) and the [University of Utrecht](https://www.uu.nl/en) in the Netherlands. The development was led by Nick Hodgskin, Joe Schoonover and Erik van Sebille, with contributions from many others, including Willi Rath, Daniel Lizarbe, and Michael Denes. | ||
|
Member
Author
There was a problem hiding this comment. Anyone else we missed? |
||
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We should refer to this version as
v4.0instead ofv4-previeworv4.0-preview. We should also capitalise package names (Xarray, Python, Zarr, Dask etc.)There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I see the point about capitalising package names (and have done so in f78bef0), but I don't entirely understand why we shouldn't call it v4.0-preview. I actually like that name; it signals clearly to our users that this is not the stable version yet. If we simply call it v4.0, then users won't know it's unstable...
Uh oh!
There was an error while loading. Please reload this page.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Because I think its quite confusing:
v4.0-previewhints that there will be av4.0release later which will be stable, when that's not the case - there may be several "preview" releases until we get to the stablev4.xversionv4.0-preview, they installv4.0. Conda nor PyPI support "preview" in the version name, "preview" is only shown when we talk about it - which I find unnecessarily confusingI think it makes a lot more sense to say that it's v4.0 - which is treated as a preview release of Parcels to get early user feedback. I think mentioning this in the blog post, the tagged release, and on import (even as a banner in the docs we could do) is good and clearly communicated.
Perhaps there is something to be said about the paper. It would be cleaner to reference the stable version as
v4.0.0, but that's just not an option if we go for this approach of doing preview releases that people get when they doconda install parcels.