Iris includes specialised handling of mesh-located data (as opposed to
grid-located data). Iris and its partner packages are
designed to make working with mesh-located data as simple as possible, with new
capabilities being added all the time. More detail is in this section and in
iris.experimental.ugrid API documentation.
This mesh support is based on the CF-UGRID Conventions; UGRID-conformant meshes + data can be loaded from a file into Iris’ data model, and meshes + data represented in Iris’ data model can be saved as a UGRID-conformant file.
- Meshes are different
Mesh-located data is fundamentally different to grid-located data. Many of Iris’ existing operations need adapting before they can work with mesh-located data, and in some cases entirely new concepts are needed. Read the detail in these pages before jumping into your own code.
- Iris’ mesh support is experimental
This is a rapidly evolving part of the codebase at time of writing (
Jan 2022), as we continually expand the operations that work with mesh data. Be prepared for breaking changes even in minor releases.
- Get involved!
We know meshes are an exciting new area for much of Earth science, so we hope there are a lot of you with new files/ideas/wishlists, and we’d love to hear more 🙂.
Read on to find out more…
The Mesh Data Model - learn why the mesh experience is so different.
Iris’ Mesh Partner Packages - meet some optional dependencies that provide powerful mesh operations.
Working with Mesh Data - experience how your workflows will look when written for mesh data.
Converting Other Mesh Formats - check out some examples of converting various mesh formats into Iris’ mesh format.