Fix typos in cell_diffusion_timescale methods.

[ali-ramadhan]

No description provided.

[codecov]

Codecov Report

Merging #557 into master will increase coverage by 3.86%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master     #557      +/-   ##
==========================================
+ Coverage   67.72%   71.59%   +3.86%     
==========================================
  Files          69       71       +2     
  Lines        1952     2341     +389     
==========================================
+ Hits         1322     1676     +354     
- Misses        630      665      +35

Impacted Files	Coverage Δ
...ure_implementations/leith_enstrophy_diffusivity.jl	98.38% <ø> (+66.12%)	⬆️
...rbulenceClosures/turbulence_closure_diagnostics.jl	100% <100%> (+80.76%)	⬆️
src/boundary_conditions.jl	69.06% <0%> (-4.97%)	⬇️
src/halo_regions.jl	83.05% <0%> (-3.07%)	⬇️
src/TimeSteppers/kernels.jl	57.8% <0%> (-2.6%)	⬇️
src/AbstractOperations/computations.jl	75.75% <0%> (-0.25%)	⬇️
src/Solvers/batched_tridiagonal_solver.jl	100% <0%> (ø)
src/logger.jl	0% <0%> (ø)
src/Oceananigans.jl	76.92% <0%> (+1.92%)	⬆️
src/OutputWriters/netcdf_output_writer.jl	89.23% <0%> (+2.69%)	⬆️
... and 9 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update a511fdd...129f822. Read the comment docs.

[glwagner]

This PR might want to add a simple test that runs these functions to ensure this doesn't continue to happen (and improve coverage), rather than only fixing the current code.

[ali-ramadhan]

@glwagner I added turbulence diagnostics tests to ensure that DiffusiveCFL is compatible with all closures.

Unfortunately it failed on TwoDimensionalLeith as it was bundled with AbstractSmagorinsky and TwoDimensionalLeith does not have a closure.Pr field.

I split it off into its own cell_diffusion_timescale where it's easy to calculate the viscous CFL condition but it seems that TwoDimensionalLeith calculates the nonlinear diffusivity on the fly and so there is no easy/cheap way to compute the diffusive CFL for each tracer.

I'm thinking of merging this PR because I need the bugfix for a script in another repo.

I will open an issue about the diffusive CFL for TwoDimensionalLeith (out of scope for this PR).