* correct

* not yet

* more changes

* remove QAB2 conservation testset

* back to this default

* fix another bug

* fix shallow water

* no need to compute these tendencies

* get it working

* all the pipelines

* should work also for multi region

* cannot use the ; syntax

* fill halo regions

* add this

* bugfix

* change stuff

* add a pressure-correction method

* go ahead

* back to runge kutta3

* do not import this

* this should work

* another bugfix

* removing one naturalsort

* include the correct file

* another bugfix

* fix more tests

* other bugfixes

* changes

* closure fields

* disambiguate

* ok this works

* fix reactant tests

* bugfix reactant tests

* distributed tests should now pass

* split into two pipelines

* do not add a reference yet

* no need to add the extra "distributed" qualifier

* let's test a mutable vertical discretization

* test also conservation in distributed grids

* not supported implicit stuff

* more conservation tests

* fix the conservation tests

* correct distributed tests

* uff

* fix multi-region

* add more tests

* another bugfix

* test like this

* put apply regionally where it matters

* reduce the number of conservation tests

* barotropic U

* U and V correction

* make sure times is on a CPU

* make sure everything is initialized correctly

* correct the signature

* use more time

* reinclude split explicit tests

* try this thing

* more changes

* Revert "more changes"

This reverts commit a822cea.

* Revert "try this thing"

This reverts commit 4c3cb39.

* change the set!

* bugfix

* some more changes

* retry the tests

* update fts

* go ahead

* remove some imports

* Update src/OrthogonalSphericalShellGrids/distributed_zipper.jl

Co-authored-by: Mosè Giordano <[email protected]>

* Update src/Oceananigans.jl

Co-authored-by: Mosè Giordano <[email protected]>

* Update Project.toml

* bugfix

* add the xflat grid and yflat grid

* remove the shows

* bugfix

* some cleanup

* start fixing up tests

* only distributed zstar tests remaining

* bugfix the nonhydrostatic model

* some updates

* this was not the cause

* add the free surface

* adding more distributed tests

* bugfix

* radius should not be 1 if z goes to -1!

* will this change stuff?

* just a check

* not sure why this would pass locally and not on buildkite

* correct imports

* other fixes

* cche current fields

* cache current fields

* remove tough tests

* deepcopy the grid to avoid mutations

* fix model

* now the tests should all pass

* bugfix

* import ZFlatGrid

* only from 0 to +1

* import KernelParameters

* continue fixing

* correct height

* do not use generic fallbacks

* some changes

* add a bunch of docstrings

* make all tests pass

* bugfix

* put back the compute_tendencies in update_state

* η -> displacement

* fix all tests

* remove the synchronize

* do not import

* compute flux bc after stepping

* remove unnecessary import

* fix single column mode

* clean imports

* minor cleanup + alphabetic order of imports

* some more cleaning up

* Fix description of barotropic pressure correction

Corrected the description of barotropic pressure gradient application in the implicit free surface method.

* fix vertical coordinate tests

* this is fixed

* uncomment tests

* comment out broken tests

* recorrect the rk-implicit

* fix all tests

* do not need update anymore

* change to quotients

* fix the cubed sphere

* another fix for the cubed sphere

* remove the quotients

* fix the ambiguities and (probably) oneAPI

* fix ambiguities

* Update src/Models/HydrostaticFreeSurfaceModels/z_star_vertical_spacing.jl

Co-authored-by: Ali Ramadhan <[email protected]>

* Update src/Models/HydrostaticFreeSurfaceModels/hydrostatic_free_surface_rk_step.jl

Co-authored-by: Ali Ramadhan <[email protected]>

* Update src/Models/interleave_communication_and_computation.jl

Co-authored-by: Ali Ramadhan <[email protected]>

* remove closure from tracer step

* address more comments on references and timestepper constructors

* avoid dividing by a 0 in the trasport velocities

* remove unused macro

* correct compute_transport_velocities

* add some documentation

* fix some remaining bugs

* Apply suggestion from @simone-silvestri

* fix tests

* find which tests fail

* catke + tke-ε only fail with QAB2

* add a shallow water AB2

* fix a couple of timesteppers

* fix checkpointer

* remove removed field

* Clean up whitespaces

* Delete src/TimeSteppers/store_tendencies.jl

* Remove inclusion of store_tendencies.jl

* fix cites

* reorder imports

* enhances some docstrings

* Update src/TimeSteppers/quasi_adams_bashforth_2.jl

Co-authored-by: Gregory L. Wagner <[email protected]>

* enhances some docstrings

* add bibtex entry for Hu et al 1996

* docstring-friendly cite

* fix signature in docstring

* clean up notation

* remove DoubleFloats

* explicit Nstages

* remove double floats

* add packages back

* add some code examples

* fix up a couple of constructors and the documentation

* add another line to the docs

* fix a bit the docs

* remove trailing whitespaces

* some grammar improvements

* fix doctests

---------

Co-authored-by: Mosè Giordano <[email protected]>
Co-authored-by: Navid C. Constantinou <[email protected]>
Co-authored-by: Ali Ramadhan <[email protected]>
Co-authored-by: Mosè Giordano <[email protected]>
Co-authored-by: Gregory L. Wagner <[email protected]>