Oceananigans.jl ocean modeling example

[hdrake]

@navidcy and @glwagner, I'd like to include an Oceananigans.jl example of an eddying double gyre in our curriculum of reactive Pluto.jl notebooks (see our current example). In the Ferrari group meeting today, we discussed making an outreach demo that would allow students to release little plastic particles in a double gyre and watch them accumulate in the convergence zones (garbage patches).

Would it make sense to use output from @navidcy's double-gyre example (CliMA/Oceananigans.jl#1085)? An alternative would be to just use a much cheaper barotropic model (i.e. nz=1) and somehow parameterize the effect that Ekman convergence has on the plastics.

What do you think?

[navidcy]

sounds great!
the example in CliMA/Oceananigans.jl#1085 is a bit expensive to run... so I'm not sure you can run it interactively on a Pluto notebook... Perhaps if we run it to equilibration then you can checkpoint it from an equilibrated state and run in the notebook for 1-2 years (with particles)? We can try!

[navidcy]

These interactive notebooks look great btw!

[hdrake]

Perhaps if we run it to equilibration then you can checkpoint it from an equilibrated state and run in the notebook for 1-2 years (with particles)?

Yeah, that could work! Alternatively, we could just load in some pre-computed outputs (e.g. 100 snapshots from the statistical steady state) and linearly interpolate (and loop) the velocities in time for the Lagrangian integrations, but maybe that is less fun.

[glwagner]

I think we should try integration offline first since I think that's most likely to be reponsive and snappy in a Pluto notebook --- what do you think @navidcy?

We can advect particles with the surface field only (whether we have Nz=1 or more levels)?