Wednesday, September 9th, 2026 (2 days ago)
We are very excited to announce that Parcels v4.0 is now available for performing Lagrangian simulations in hydrodynamic flows. Parcels v4 is a complete redesign of the Parcels code and is designed to be more flexible and extensible than previous versions.
With this release - Parcels is a pure Python package that leverages the full power of the Pangeo ecosystem 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 easier to implement.
Key improvements include:
We have also changed the output format to Parquet, which is a tabular format more suited for writing and reading trajectory data.
See the full migration guide from Parcels v3 to v4 here.
Note that Parcels v4.0 is an early release to gather user feedback and should not be considered stable. We welcome feedback from the community, and encourage users to try out the new version and report any issues or suggestions or connect with us on our Zulip CLAM community.
Also note that the performance of Parcels v4 is currently slower than Parcels v3 for large (> ~100,000) numbers of particles. We are working on improving performance and will release updates as we make progress.
Once we've incorporated this feedback, improved performance, and implemented a few outstanding features, we aim to release a stable version of Parcels.
Funded by the WarmWorld ELPHE project 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”, the development of Parcels v4 started in September 2025 and was a collaboration between the GEOMAR Helmholtz Centre for Ocean Research Kiel, FluidNumerics and the University of Utrecht 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, Reint Fischer, Wyatt Sieminski, and Michael Denes.