## Oceananigans v0.76.7

[Diff since v0.76.6](CliMA/Oceananigans.jl@v0.76.6...v0.76.7)


**Closed issues:**
- Need actual doctstring for `SpecifiedTimes` (CliMA#2608)
- Need to document the available solver method options for ImplicitFreeSurface solvers (CliMA#2620)
- Need to check for constant depth *before* we use the `FFTBasedPoissonSolver` in `HydrostaticFreeSurfaceModel`? (CliMA#2626)
- Rename `Poisson Solvers` section in Docs? (CliMA#2637)
- Is `calc_∇²!` redundant in view of `compute_∇²!`? (CliMA#2638)

**Merged pull requests:**
- Partial cell immersed boundaries (CliMA#2306) (@francispoulin)
- Implements a multigrid solver for Poisson and implicit free surface equations on CPU (CliMA#2396) (@glwagner)
- Fields in viscosity operators (CliMA#2580) (@simone-silvestri)
- Add docstring for `SpecifiedTimes` (CliMA#2617) (@navidcy)
- Update package's dependencies (CliMA#2619) (@navidcy)
- Update `ImplicitFreeSurface` & `HeptadiagonalIterativeSolver` docstrings (CliMA#2621) (@navidcy)
- Add empty PR template (CliMA#2624) (@navidcy)
- Docs library cleanup (CliMA#2627) (@navidcy)
- Some updates in validation scripts: `AveragedField()` -> `Field(Average())` (CliMA#2630) (@navidcy)
- Fix typo in Docs/Buoyancy models (CliMA#2633) (@navidcy)
- Small bugfix for `ContinuousBoundaryFunction` (CliMA#2634) (@simone-silvestri)
- Use predefined `compute_∇²!` instead of defining new `calc_∇²!` in `test_matrix_poisson_solver.jl` (CliMA#2640) (@navidcy)
- Rename `Docs/Numerical Implementation/Poisson solvers` section -> `Elliptic solvers` (CliMA#2641) (@navidcy)
- `VectorInvariant` as a default for `HydrostaticFreeSurfaceModel` with `LatitudeLongitudeGrid` (CliMA#2644) (@simone-silvestri)
- Minor changes in Langmuir and Ocean wind mixing examples (CliMA#2646) (@tomchor)
- Small change to 1D diffusion example (CliMA#2648) (@tomchor)

## Oceananigans v0.76.6

[Diff since v0.76.5](CliMA/Oceananigans.jl@v0.76.5...v0.76.6)


**Closed issues:**
- CUDA error with immersed boundary when running on GPU (CliMA#2479)
- Odd behavior using `ImmersedBoundary` while specifying scratch data for `Field` computation (CliMA#2581)
- `closure` not working in `ShallowWaterModel` (CliMA#2606)
- Can we have advection like terms in the forcing? (CliMA#2609)

**Merged pull requests:**
- Near global one degree setup (CliMA#2463) (@glwagner)
- Improve support for immersed boundaries with `Flat` (CliMA#2532) (@glwagner)
- CompatHelper: bump compat for DocStringExtensions to 0.9, (keep existing compat) (CliMA#2578) (@github-actions[bot])
- `compute_at!` for ConditionalOperation (CliMA#2582) (@glwagner)
- Update `near_global_quarter_degree.jl` (CliMA#2593) (@simone-silvestri)
- Update boundary_conditions.md (CliMA#2598) (@francispoulin)
- Updates for Docs/Physics section (CliMA#2599) (@navidcy)
- Drop SafeTestset from package dependency (CliMA#2601) (@navidcy)
- Fix `ShallowWaterModel` closure for tracers in the vector invariant form (CliMA#2607) (@francispoulin)
- Update `ShallowWaterModel` to use areas (CliMA#2612) (@francispoulin)
- Bump up PencilFFTs (CliMA#2613) (@navidcy)
- Adds back a fig in Docs/Model/Grids + example with `Flat` topology (CliMA#2614) (@navidcy)

## Oceananigans v0.76.5

[Diff since v0.76.4](CliMA/Oceananigans.jl@v0.76.4...v0.76.5)


**Closed issues:**
- Can't we have a 2D simulation on the sphere? (CliMA#2372)
- Does non hydrostatic model impose boundary conditions based on topology? (CliMA#2594)
- Unnecessary GPU check? (CliMA#2595)

**Merged pull requests:**
- Vector Invariant Formulation of the `ShallowWaterModel` (CliMA#2522) (@simone-silvestri)
- Check for CUDA-enabled device on `grid` construction rather than in model constructors (CliMA#2596) (@navidcy)

## Oceananigans v0.76.4

[Diff since v0.76.3](CliMA/Oceananigans.jl@v0.76.3...v0.76.4)


**Closed issues:**
- Error using immersed boundary on GPUs with buoyancy tracer (CliMA#2558)
- Type Error using ImmersedBoundaries on GPU (CliMA#2563)
- Do we want a `HorizontalDivergenceField`? (CliMA#2575)
- using `∂x` in a forcing function (CliMA#2586)
- Input validation for stretched grid dimensions breaks `FieldTimeSeries` (CliMA#2588)

**Merged pull requests:**
- Synchronous intra-device copy (CliMA#2547) (@simone-silvestri)
- Update PencilArrays (CliMA#2552) (@navidcy)
- Convert `two_dimensional_turbulence.jl` and `internal_wave.jl` literated examples in docs to use CairoMakie (CliMA#2553) (@navidcy)
- Adds tracer-specific diffusivity subsection to `model_setup` (CliMA#2554) (@tomchor)
- Convert more examples to plot with Makie (CliMA#2556) (@navidcy)
- Maximum-principle-satisfying WENO scheme (CliMA#2557) (@simone-silvestri)
- Regularize Number boundary conditions to eltype(grid) (CliMA#2559) (@glwagner)
- Completely trivial update to a comment about continguous CuArrays (CliMA#2561) (@glwagner)
- Eliminate Plots as dependency in Docs; use Makie (CliMA#2562) (@navidcy)
- Adapt for `ImmersedBoundaryCondition` (CliMA#2564) (@simone-silvestri)
- Adds `with_halos` to `NetCDFOutputWriter`'s signature call in docstring (CliMA#2566) (@tomchor)
- One more one(grid) (CliMA#2567) (@glwagner)
- Implement lastindex for FieldTimeSeries (CliMA#2568) (@glwagner)
- Fix inconsistent behaviour in `test_matrix_poisson_solver.jl` (CliMA#2569) (@simone-silvestri)
- update `near_global_quarter_degree.jl` (CliMA#2570) (@francispoulin)
- Fixes small typo in Langmuir turbulence example (CliMA#2571) (@tomchor)
- Fix matrix solver test take 2 (CliMA#2572) (@simone-silvestri)
- small bugfix in `Integral` construction (CliMA#2573) (@simone-silvestri)
- Update dependencies (CliMA#2574) (@navidcy)
- Beautify examples a bit more (CliMA#2576) (@glwagner)
- To enable discrete functions on "ImmersedBoundaryConditions" (CliMA#2577) (@simone-silvestri)
- `PreconditionedConjugateGradientSolver`: fix docstring and `show()` method (CliMA#2583) (@navidcy)
- Fix admonition (CliMA#2585) (@navidcy)
- Bug fix in conditional `advective_tracer_flux` (CliMA#2587) (@simone-silvestri)
- Fix fatal interaction between FieldTimeSeries construction and input validation (CliMA#2589) (@glwagner)

## Oceananigans v0.76.3

[Diff since v0.76.2](CliMA/Oceananigans.jl@v0.76.2...v0.76.3)


**Closed issues:**
- Should we check the size of grids in stretched directions? (CliMA#2548)

**Merged pull requests:**
- Throws error when size of dimension is mismatched when constructing RectilinearGrid (CliMA#2550) (@tomchor)
- Fixes bug in IBM with Value or Gradient boundary conditions (CliMA#2551) (@tomchor)

## Oceananigans v0.76.2

[Diff since v0.76.1](CliMA/Oceananigans.jl@v0.76.1...v0.76.2)


**Closed issues:**
- `NetCDF`: filename vs filepath (CliMA#2541)
- error in `baroclinic_adjustment.jl` (CliMA#2543)

**Merged pull requests:**
- Update CATKE parameters for new implicit time-stepping (CliMA#2461) (@glwagner)
- Permute PencilFFT temporary array dimensions (CliMA#2513) (@glwagner)
- Cosmetic improvement to checkpointer docs (CliMA#2531) (@glwagner)
- Support two-dimensional distributed FFT-based Poisson solves (CliMA#2536) (@glwagner)
- Make JLD2 look before leaping (CliMA#2537) (@glwagner)
- CompatHelper: add new compat entry for PencilArrays at version 0.16, (keep existing compat) (CliMA#2539) (@github-actions[bot])
- Update PencilArrays and PencilFFTs (CliMA#2544) (@github-actions[bot])
- Fixes `WENO5(grid=grid)` for immersed boundary grids (CliMA#2549) (@tomchor)

## Oceananigans v0.76.1

[Diff since v0.76.0](CliMA/Oceananigans.jl@v0.76.0...v0.76.1)


**Closed issues:**
- TEOS10 Equation of State: Depth vs Height (CliMA#2528)

**Merged pull requests:**
- Update dependency packages (CliMA#2447) (@navidcy)
- Update compat for `SeawaterPolynomials` (CliMA#2529) (@glwagner)

## Oceananigans v0.76.0

[Diff since v0.75.3](CliMA/Oceananigans.jl@v0.75.3...v0.76.0)


**Closed issues:**
- Shared memory for GPU kernels (CliMA#442)
- `NetCDFOutputWriter` leaves file open if the constructor errors (CliMA#555)
- `InterpolatedField` for Lagrangian particle tracking, output on custom grids, and restoring from checkpoint to a new grid. (CliMA#975)
- Different advection schemes for velocities and tracers (CliMA#1015)
- `Terrain-following` coordinates anyone? (CliMA#1549)
- `Connected` topologies (CliMA#1582)
- Error from benchmark/test distributed nonhydrostatic model scripts (CliMA#2347)
- Why do we need a `hack_cosd`/`hack_sind`? (CliMA#2371)
- Oceananigans Port to AMDGPU (CliMA#2493)

**Merged pull requests:**
- Multi-region module for MPI-less multi-GPU computation  (CliMA#2253) (@simone-silvestri)
- VerticallyImplicitTimeDiscretization for symmetric isoneutral diffusion (CliMA#2487) (@glwagner)
- AMDGPU Branch Initial Pull Request (CliMA#2494) (@matinraayai)
- Remove spurious extra calculate_diffusivities! (CliMA#2498) (@glwagner)
- Fix minor error in `internal_wave.jl` (CliMA#2500) (@glwagner)
- Update PencilFFTs and GPUArrays and fix DistributedFFTBasedPoissonSolver (CliMA#2502) (@glwagner)
- Update `JLD2OutputWriter` calls (CliMA#2504) (@navidcy)
- Fix url for section about papers citing Oceananigans (CliMA#2505) (@navidcy)
- Update CubedSpheres (CliMA#2506) (@jm-c)
- Remove vertical advection from vector invariant WENO (CliMA#2510) (@simone-silvestri)
- README tweak (CliMA#2514) (@glwagner)
- Drop direct dependency to GPUArrays and BenchmarkTools (CliMA#2518) (@navidcy)
- Add tests for computations on slices (CliMA#2521) (@glwagner)
- OutputWriters for MultiRegion (CliMA#2524) (@simone-silvestri)

## Oceananigans v0.75.3

[Diff since v0.75.2](CliMA/Oceananigans.jl@v0.75.2...v0.75.3)


**Closed issues:**
- Problem installing and using Oceananigans (CliMA#2480)

**Merged pull requests:**
- Implement the actually-Cox (aka the anti-anti-Cox) stencil for GM (CliMA#2481) (@glwagner)
- Remove redundant definition of `diffusive_flux_z` for divergence damping (CliMA#2482) (@glwagner)
- Pin MPICH_jll at 4.0.1 (CliMA#2486) (@glwagner)
- Fix docs url typo in `README` (CliMA#2491) (@navidcy)
- Biharmonic divergence damping (CliMA#2495) (@simone-silvestri)
- Drop MPICH_jll direct dependency (CliMA#2496) (@navidcy)

## Oceananigans v0.75.2

[Diff since v0.75.1](CliMA/Oceananigans.jl@v0.75.1...v0.75.2)


**Closed issues:**
- Minor problem with `closure` in `show(io, :: NonhydrostaticModel)`? (CliMA#2443)
- Smsm (CliMA#2466)
- Smsm (CliMA#2468)

**Merged pull requests:**
- Useful regression benchmark (CliMA#1214) (@ali-ramadhan)
- Support boundary conditions on immersed boundaries (CliMA#2437) (@glwagner)
- Adds HorizontalDivergenceFormulation for scalar diffusions (CliMA#2440) (@glwagner)
- Few beautifications in docstrings (CliMA#2458) (@navidcy)
- Few minor beauty fixes (CliMA#2459) (@navidcy)
- Better diffusivity closure `show()` (CliMA#2460) (@navidcy)