Adds ORCA2 tripolar grid and changed ORCAGrid to ORCATripolarGrid by jagoosw · Pull Request #153 · NumericalEarth/NumericalEarth.jl

jagoosw · 2026-04-02T19:58:22Z

This PR adds ORCA2 grids to data wrangling and makes some changes to grid construction to get it working. It also renamed ORCAGrid to ORCATripolarGrid, adds a "coarse" example on ORCA2, and adds the option to remove closed basins int he grid constructor.

ORCA2 is strange because it has regional refinement in the Mediterranean, Red, Black and Caspian Seas, as well as at the equator and south pole.

The underlying grid:

Nz = 31
z = ExponentialDiscretization(31, -5000, 0; scale = 1100)
grid = ORCAGrid(z = z.faces, Nz, dataset = ORCA2())

I had to add Tar and CodecZlib to the dependancies because the Zenodo I found with the grid has them in a tarball.

@simone-silvestri @glwagner

jagoosw · 2026-04-02T20:00:10Z

+        @info "Downloading ORCA2 data: $(metadatum.name) to $(metadatum.dir)..."
+        Downloads.download(fileurl, fileroute*"ORCA2L75.tar.gz"; progress=download_progress)
+        open(GzipDecompressorStream, fileroute*"ORCA2L75.tar.gz") do io
+            Tar.extract(io, fileroute*"/ORCA2L75")


Tar.extract doesn't like the destination fileroute so the files actually live at fileroute*"/ORCA2L75/ORCA2L75"

codecov · 2026-04-02T20:20:10Z

Codecov Report

❌ Patch coverage is 83.33333% with 5 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
src/DataWrangling/ORCA/ORCA.jl	80.95%	4 Missing ⚠️
src/Bathymetry/orca_tripolar_grid.jl	88.88%	1 Missing ⚠️

📢 Thoughts on this report? Let us know!

glwagner · 2026-04-02T21:57:00Z

Is this a tripolar grid? Does it have a zipper bc? Can we call it ORCATripolarGrid?

glwagner · 2026-04-02T21:57:45Z

I had to add Tar and CodecZlib to the dependancies because the Zenodo I found with the grid has them in a tarball.

@giordano is this correct or should we use an extension

@jagoosw one possibility is to put this in a new package (we can put it in NumericalEarth) and then add an extension for the deps

giordano · 2026-04-02T22:33:57Z

@giordano is this correct or should we use an extension

I think this is correct: if this was in an extension, users would need to know in advance what kind of files they're downloading and explicitly load Tar/CodecZlib just to be safe (I'd assume at least Tar is already loaded via other dependencies anyway), which doesn't seem to be very user-friendly. Also, pkgextensions usually are meant to extend methods on types defined in the packages they glue, but the downloaded file isn't a different type depending on its format, so there wouldn't be many types for which to extend methods.

jagoosw · 2026-04-03T13:47:50Z

@jagoosw one possibility is to put this in a new package (we can put it in NumericalEarth) and then add an extension for the deps

Yeah its very similar to ORCA1 which is already in the file I added to, they just have some regional refinement too.

https://www.nemo-ocean.eu/doc/node108.html#Tab_orca_zgr

jagoosw · 2026-04-03T13:48:38Z

I think this is correct: if this was in an extension, users would need to know in advance what kind of files they're downloading and explicitly load Tar/CodecZlib just to be safe (I'd assume at least Tar is already loaded via other dependencies anyway), which doesn't seem to be very user-friendly. Also, pkgextensions usually are meant to extend methods on types defined in the packages they glue, but the downloaded file isn't a different type depending on its format, so there wouldn't be many types for which to extend methods.

I think you're right that Tar was already loaded

jagoosw · 2026-04-03T19:56:44Z

When I try to run the example I get NaNs on the first time step, it seems ORCA1 does the same too? (@simone-silvestri)

(The example I added is a copy of the one degree example, I don't know if we want it in the examples folder in this PR but there doesn't seem to be a validation folder)

glwagner · 2026-04-03T19:58:27Z

 z = ExponentialDiscretization(Nz, -depth, 0; scale = depth/4, mutable = true)

-grid = ORCAGri(arch, z = z.face, Nz, dataset = ORCA2())
+grid = ORCAGrid(arch; z, Nz, dataset = ORCA2(), halo = (5, 5, 4))


Can we use ORCATripolarGrid ?

glwagner

ORCATripolarGrid?

simone-silvestri · 2026-04-07T12:40:11Z

There were a couple of bugs that have been solved, now I can run an ORCA1 with 30minutes quite stably

jagoosw · 2026-04-24T14:30:12Z

I'm just sorting out the conflicts and everything is straight forward except the tests. I had modified the tests to do do all the same tests on the ORCA2 grid as the ORCA1 grid, but the incoming changes for ORCA12 don't test everything like metric consistency. Thinking about it now we probably don't need todo all the tests on all the grids because if the constructor works for one it should work for all of them? Should I make the tests just do basic tests on ORCA2 like ORCA12, or should I extend to include ORCA12 in all the tests, or do all the tests on 1 and 2 but not 12? @simone-silvestri @taimoorsohail

simone-silvestri · 2026-04-24T14:40:01Z

I think they are quite simple tests so worth to add also ORCA12 to the more in-depth testing

jagoosw · 2026-04-24T14:48:32Z

Okay I added them and ORCA12 is failing the metric consistency tests:

ORCATripolarGrid (ORCA12) metric consistency: Test Failed at /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102
  Expression: all((x->begin
                #= /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 =#
                x > 0
            end), interior) == true
   Evaluated: false == true

Stacktrace:
 [1] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:672 [inlined]
 [2] macro expansion
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 [inlined]
 [3] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:1577 [inlined]
 [4] top-level scope
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:80
ORCATripolarGrid (ORCA12) metric consistency: Test Failed at /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102
  Expression: all((x->begin
                #= /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 =#
                x > 0
            end), interior) == true
   Evaluated: false == true

Stacktrace:
 [1] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:672 [inlined]
 [2] macro expansion
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 [inlined]
 [3] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:1577 [inlined]
 [4] top-level scope
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:80
ORCATripolarGrid (ORCA12) metric consistency: Test Failed at /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102
  Expression: all((x->begin
                #= /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 =#
                x > 0
            end), interior) == true
   Evaluated: false == true

Stacktrace:
 [1] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:672 [inlined]
 [2] macro expansion
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 [inlined]
 [3] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:1577 [inlined]
 [4] top-level scope
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:80
ORCATripolarGrid (ORCA12) metric consistency: Test Failed at /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102
  Expression: all((x->begin
                #= /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 =#
                x > 0
            end), interior) == true
   Evaluated: false == true

Stacktrace:
 [1] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:672 [inlined]
 [2] macro expansion
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 [inlined]
 [3] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:1577 [inlined]
 [4] top-level scope
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:80
ORCATripolarGrid (ORCA12) metric consistency: Test Failed at /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102
  Expression: all((x->begin
                #= /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 =#
                x > 0
            end), interior) == true
   Evaluated: false == true

Stacktrace:
 [1] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:672 [inlined]
 [2] macro expansion
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 [inlined]
 [3] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:1577 [inlined]
 [4] top-level scope
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:80
ORCATripolarGrid (ORCA12) metric consistency: Test Failed at /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102
  Expression: all((x->begin
                #= /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 =#
                x > 0
            end), interior) == true
   Evaluated: false == true

Stacktrace:
 [1] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:672 [inlined]
 [2] macro expansion
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 [inlined]
 [3] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:1577 [inlined]
 [4] top-level scope
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:80
ORCATripolarGrid (ORCA12) metric consistency: Test Failed at /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102
  Expression: all((x->begin
                #= /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 =#
                x > 0
            end), interior) == true
   Evaluated: false == true

Stacktrace:
 [1] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:672 [inlined]
 [2] macro expansion
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 [inlined]
 [3] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:1577 [inlined]
 [4] top-level scope
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:80
ORCATripolarGrid (ORCA12) metric consistency: Test Failed at /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102
  Expression: all((x->begin
                #= /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 =#
                x > 0
            end), interior) == true
   Evaluated: false == true

Stacktrace:
 [1] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:672 [inlined]
 [2] macro expansion
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 [inlined]
 [3] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:1577 [inlined]
 [4] top-level scope
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:80
ORCATripolarGrid (ORCA12) metric consistency: Test Failed at /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102
  Expression: all((x->begin
                #= /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 =#
                x > 0
            end), interior) == true
   Evaluated: false == true

Stacktrace:
 [1] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:672 [inlined]
 [2] macro expansion
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 [inlined]
 [3] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:1577 [inlined]
 [4] top-level scope
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:80
ORCATripolarGrid (ORCA12) metric consistency: Test Failed at /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102
  Expression: all((x->begin
                #= /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 =#
                x > 0
            end), interior) == true
   Evaluated: false == true

Stacktrace:
 [1] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:672 [inlined]
 [2] macro expansion
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 [inlined]
 [3] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:1577 [inlined]
 [4] top-level scope
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:80
ORCATripolarGrid (ORCA12) metric consistency: Test Failed at /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102
  Expression: all((x->begin
                #= /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 =#
                x > 0
            end), interior) == true
   Evaluated: false == true

Stacktrace:
 [1] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:672 [inlined]
 [2] macro expansion
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 [inlined]
 [3] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:1577 [inlined]
 [4] top-level scope
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:80
ORCATripolarGrid (ORCA12) metric consistency: Test Failed at /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102
  Expression: all((x->begin
                #= /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 =#
                x > 0
            end), interior) == true
   Evaluated: false == true

Stacktrace:
 [1] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:672 [inlined]
 [2] macro expansion
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:102 [inlined]
 [3] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:1577 [inlined]
 [4] top-level scope
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:80
ORCATripolarGrid (ORCA12) metric consistency: Test Failed at /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:121
  Expression: n_west / Nx > 0.95
   Evaluated: 0.5002313743637204 > 0.95

Stacktrace:
 [1] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:672 [inlined]
 [2] macro expansion
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:121 [inlined]
 [3] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:1577 [inlined]
 [4] top-level scope
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:80
ORCATripolarGrid (ORCA12) metric consistency: Test Failed at /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:129
  Expression: nsouth / length(φC) > 0.95
   Evaluated: 0.4337215751525236 > 0.95

Stacktrace:
 [1] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:672 [inlined]
 [2] macro expansion
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:129 [inlined]
 [3] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:1577 [inlined]
 [4] top-level scope
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:80
ORCATripolarGrid (ORCA12) metric consistency: Test Failed at /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:140
  Expression: boundary_jump < 10interior_variation + 1.0e-10
   Evaluated: NaN < NaN

Stacktrace:
 [1] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:672 [inlined]
 [2] macro expansion
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:140 [inlined]
 [3] macro expansion
   @ ~/.julia/juliaup/julia-1.10.10+0.aarch64.apple.darwin14/share/julia/stdlib/v1.10/Test/src/Test.jl:1577 [inlined]
 [4] top-level scope
   @ ~/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:80
Test Summary:                                | Pass  Fail  Total   Time
ORCATripolarGrid (ORCA12) metric consistency |   20    15     35  17.0s
ERROR: LoadError: Some tests did not pass: 20 passed, 15 failed, 0 errored, 0 broken.
in expression starting at /Users/jago/Documents/Projects/NumericalEarth.jl/test/test_orca_tripolar_grid.jl:17

@taimoorsohail have you managed to run a model using the ORCA12 grid?

simone-silvestri · 2026-04-24T14:49:25Z

Ok then I guess and issue that tracks the failure is more suited 😅 So we can merge this PR and deal with ORCA12 in another PR

jagoosw · 2026-04-24T14:52:34Z

To summarise I think the tests are telling us that:

ORCA12 has interior points with grid spacing or areas <= 0
Lots of face points are the "wrong" side of the center points
Theres something wrong with the periodic boundary

jagoosw · 2026-04-24T14:52:54Z

Ok then I guess and issue that tracks the failure is more suited 😅 So we can merge this PR and deal with ORCA12 in another PR

~okay I'll mark the tests as broken ~ I've commented out ORCA12 from the tests because it was easier than marking every test as broken

jagoosw · 2026-04-24T14:55:37Z

I think this is ready for review now. @simone-silvestri do you think I should reduce the stop time for the example?

simone-silvestri · 2026-04-24T15:10:59Z

+# Now we are ready to build the coupled ocean--sea ice model and bring everything
+# together into a `simulation`.
+
+# With Runge-Kutta 3rd order time-stepping we can safely use a timestep of 20 minutes.


Suggested change

# With Runge-Kutta 3rd order time-stepping we can safely use a timestep of 20 minutes.

# With Runge-Kutta 3rd order time-stepping we can safely use a timestep of 90 minutes.

simone-silvestri · 2026-04-27T10:45:16Z

+vertical_mixing = CATKEVerticalDiffusivity(VerticallyImplicitTimeDiscretization(); 
+                                           maximum_tracer_diffusivity = 1,
+                                           maximum_tke_diffusivity = 1,
+                                           maximum_viscosity = 1)


we cannot use the default_ocean_closure?

I think I was getting instability when I didn't have it capped, but I'll rerun and check

I don't think we should have any examples that don't use the default --- we should fix the default instead.

simone-silvestri · 2026-04-27T10:46:03Z

+tracer_advection   = WENO(order=5)
+
+dates = DateTime(1993, 1, 1) : Month(1) : DateTime(1993, 11, 1)
+mask = LinearlyTaperedPolarMask(southern=(-80, -70), northern=(70, 90), z=(-100, 0))


do we need to restore under sea ice?

Not sure, should I try without the mask?

yeah I think so.

simone-silvestri · 2026-04-27T10:47:09Z

+        @test size(bathy) == (Nx0, Ny0)
+        @test maximum(bathy) > 5000  # Deep ocean
+    end
+end


Suggested change

end

end

Co-authored-by: Simone Silvestri <silvestri.simone0@gmail.com>

jagoosw · 2026-04-27T14:38:05Z

I think I've resolved the test conflicts but because I've renamed the test file it didn't show me the diffs automatically, is it just the new bit to predownload the datasets?

simone-silvestri · 2026-04-29T08:43:14Z

I think the last bit before merging is to add these to the CATKE defaults:

maximum_tracer_diffusivity = 1
maximum_tke_diffusivity = 1
maximum_viscosity = 1

(If they are truly needed). I don't think it will damage the accuracy of the simulation, given that 1 $m^2/s$ is already quite a large diffusivity but I defer to @glwagner to assess if this is a correct move.

simone-silvestri · 2026-04-29T08:45:36Z

I think I've resolved the test conflicts but because I've renamed the test file it didn't show me the diffs automatically, is it just the new bit to predownload the datasets?

you mean for test_orca_grid? I think so but you can confirm by diffing what it is main now

jagoosw · 2026-04-29T13:32:50Z

I just checked and when I run without the limits I got NaNs. If I recall correctly, when I originally investigated this it was coming from a storm. I think maybe its a problem that is manifesting on coarse grids because if e.g. we interpolate the atmosphere fields and the coarse point happens to be near a very high wind speed point then that high wind speed is applied over the whole cell injecting a lot more momentum than on a more refined grid that resolves the variability in the atmosphere. I don't know if there are other strategies for downscaling the atmosphere some kind of averaging that people use?

glwagner · 2026-04-29T14:28:40Z

My main concern is that the limiters are unphysical. A diffusivity that large can occur for winds of 33 m/s and a 1 m mixing length, which seems possible in extreme conditions for sure.

The question is why 1 m^2 / s lead to blow up.

jagoosw · 2026-04-29T15:49:04Z

My main concern is that the limiters are unphysical. A diffusivity that large can occur for winds of 33 m/s and a 1 m mixing length, which seems possible in extreme conditions for sure.

The question is why 1 m^2 / s lead to blow up.

1m^2/s was an arbitrary choice, I can try running with it higher to see what the stable maximum is. I'm not sure exactly how the instability is arising, I think there can be a feedback with the bulk formula fluxs if the fluxs become large which could happen if theres a deep mixing event and e.g. the SST suddenly drops?

jagoosw and others added 13 commits March 30, 2026 21:07

Update ORCA.jl

ad68434

working

b2b9359

maybe now

b3136d8

maybe now

10c51bf

oops

998a1f9

remove closed basins

0341c29

remove closed basins

7dc3e0b

weird order required

83bb1ff

still doesnt work

b802a28

Removed random attampts to fix things

b2edac5

fix bathymetry

35c8dc5

Merge remote-tracking branch 'origin' into jsw/orca2

92c6dc0

export ORCA2

b694ff8

jagoosw commented Apr 2, 2026

View reviewed changes

attempt at validation

d60df12

jagoosw force-pushed the jsw/orca2 branch from 9f4a0a6 to d60df12 Compare April 3, 2026 13:52

jagoosw added 3 commits April 3, 2026 15:13

Wrap bathymetry in field to make it work on GPU

51705f4

added tests

7376478

changed halos

676d3e1

glwagner reviewed Apr 3, 2026

View reviewed changes

Merge branch 'main' into jsw/orca2

99c763f

resolve conflicts

da2495c

remove ORCA12 from tests

ffc8102

simone-silvestri reviewed Apr 24, 2026

View reviewed changes

jagoosw changed the title ~~Adds ORCA2TripolarGrid and changed ORCAGrid to ORCATripolarGrid~~ Adds ORCA2 tripolar grid and changed ORCAGrid to ORCATripolarGrid Apr 24, 2026

Update orca2.jl

2a94e87

simone-silvestri reviewed Apr 27, 2026

View reviewed changes

Comment thread examples/orca2.jl Outdated

simone-silvestri reviewed Apr 27, 2026

View reviewed changes

Comment thread examples/orca2.jl Outdated

simone-silvestri reviewed Apr 27, 2026

View reviewed changes

simone-silvestri approved these changes Apr 27, 2026

View reviewed changes

jagoosw and others added 2 commits April 27, 2026 15:17

Apply suggestions from code review

a3f0c77

Co-authored-by: Simone Silvestri <silvestri.simone0@gmail.com>

resolve conflicts

68b35c6

jagoosw added 2 commits April 27, 2026 18:09

missed test line

d4f5abd

restoring under ice

b588697

Merge branch 'main' into jsw/orca2

7eab87c

Merge remote-tracking branch 'origin' into jsw/orca2

f3f81b2

	# With Runge-Kutta 3rd order time-stepping we can safely use a timestep of 20 minutes.
	# With Runge-Kutta 3rd order time-stepping we can safely use a timestep of 90 minutes.

Conversation

jagoosw commented Apr 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

codecov Bot commented Apr 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

glwagner commented Apr 2, 2026

Uh oh!

glwagner commented Apr 2, 2026

Uh oh!

giordano commented Apr 2, 2026

Uh oh!

jagoosw commented Apr 3, 2026

Uh oh!

jagoosw commented Apr 3, 2026

Uh oh!

jagoosw commented Apr 3, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

glwagner left a comment

Choose a reason for hiding this comment

Uh oh!

simone-silvestri commented Apr 7, 2026

Uh oh!

jagoosw commented Apr 24, 2026

Uh oh!

simone-silvestri commented Apr 24, 2026

Uh oh!

jagoosw commented Apr 24, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

simone-silvestri commented Apr 24, 2026

Uh oh!

jagoosw commented Apr 24, 2026

Uh oh!

jagoosw commented Apr 24, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jagoosw commented Apr 24, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jagoosw commented Apr 27, 2026

Uh oh!

simone-silvestri commented Apr 29, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

simone-silvestri commented Apr 29, 2026

Uh oh!

jagoosw commented Apr 29, 2026

Uh oh!

glwagner commented Apr 29, 2026

Uh oh!

jagoosw commented Apr 29, 2026

Uh oh!

Reviewers

Assignees

Labels

jagoosw commented Apr 2, 2026 •

edited

Loading

codecov Bot commented Apr 2, 2026 •

edited

Loading

jagoosw commented Apr 3, 2026 •

edited

Loading

jagoosw commented Apr 24, 2026 •

edited

Loading

jagoosw commented Apr 24, 2026 •

edited

Loading

simone-silvestri commented Apr 29, 2026 •

edited

Loading