resting_stratified_bumpy_ocean.jl not running

[francispoulin]

I tried running the above in validation and it no longer works. I get the following error.

Can someone suggest how this should be updated? I am happy to try it out.

ERROR: LoadError: ArgumentError: The dimensions of the immersed boundary (1, 134, 1) do not match the grid size (1, 134)

[glwagner]

Link to file?

[francispoulin]

Sorry for not doing this earlier but here goes.

https://github.com/CliMA/Oceananigans.jl/blob/main/validation/immersed_boundaries/resting_stratified_bumpy_ocean.jl

[glwagner]

We can use functions with immersed boundaries now, so this might work:

immersed_boundaries = [
                       PartialCellBottom(seamount;
                                         minimum_fractional_Δz),
                       GridFittedBottom(seamount)
                      ]

and you don't need seamount_field.

[glwagner]

This would probably work too:

immersed_boundaries = [
                       PartialCellBottom(interior(seamount_field, :, :, 1);
                                         minimum_fractional_Δz),
                       GridFittedBottom(interior(seamount_field, :, :, 1))
                      ]

[francispoulin]

Thanks!

I can confirm that the first one works. Since this method doesn't require defining seamount_field, that seems the better way to go.

I'll create a PR right now with this update.

[glwagner]

great, thanks

[francispoulin]

One more question @glwagner on this.

I noticed that if I specified the advection schemes to be WENO, then an error occurs because the immersed boundary is (1, 136) and it does not match the grid size (1, 138). I tried increasin the halo to 4 in all directions and that didn't help.

I am pretty sure I am doing something wrong but I don't know what am I missing here. Any suggestions?

tracer_advection = WENO(grid = grid),
momentum_advection = WENO(grid = grid),

[glwagner]

This is the bug in #2988 ... try

tracer_advection = WENO(grid.underlying_grid)
momentum_advection = WENO(grid.underlying_grid)

@simone-silvestri

[glwagner]

sorry about that, it's very annoying.

[francispoulin]

Thanks for the suggestion and will try it out soon. But now to catch a plane Boston. Too bad I won't get to visit MIT as I go straight to WHOI for a workshop.