Creating a Nested Grid Tutorial

[erikvansebille]

This PR creates a tutorial for how to do nesting in Parcels v4, using a combination of unstructured (for the nest localisation) and structured (for the hydrodynamic data) Fields.

It is different from #2436 in that it uses idealised flow fields, so can be run in CI

[reint-fischer]

Very cool notebook, and so exciting to see this feature in v4! My comments are mostly considering the cases other people will use this feature for.

• When showing the velocity fields in rectangles with the Grid polygons in red, maybe add a note (admonition) discussing the case when the fields come from a nested model: the velocity fields would then be defined for the polygons only, and you would only need to pass the bounding coordinates?
• The triangulation function seems to be the core function that people will use (copy?) to set up their own nested fields. I was thinking about how to make this robust, maybe we can at least link to the documentation of the triangle function used: https://rufat.be/triangle/delaunay.html
• The AdvectionEE_NestedGrids will also be used a lot I assume. The copying of the variables is not so bad, but maybe we should add the errorcode TODO as a box to be checked to Refactor particleset attributes and particle data for vectorized kernels #2143? Alternatively, is there not a way to pass particles[mask] as a fifth argument?

After writing this I also think it is okay to fix some of these challenges once people start implementing their own nested setups in v4 and come with feedback.

[erikvansebille]

I implemented the first two suggestions in 3d1be3e. The third one will indeed be fixed/cleaned when we solve #2143