fix clock argument in TriadIsopycnalSkewSymmetricDiffusivity's explicit_R₃₃_∂z_c function

[xkykai]

Closes #4779

[glwagner]

Can you add a test that would catch this?

[xkykai]

another missing method bug was caught during test writing:

Oceananigans.jl/src/TurbulenceClosures/abstract_scalar_diffusivity_closure.jl

Line 315 in f9cba31

@inline κᶜᶜᶜ(i, j, k, grid, loc, κ::Number, clk, fields) = κ

(funnily enough I remembered catching this bug a few weeks ago and I thought I placed a fix but I can't find where that fix was)

and here are some missing function arguments too in the diffusivity fields:

Oceananigans.jl/src/TurbulenceClosures/turbulence_closure_implementations/isopycnal_skew_symmetric_diffusivity_with_triads.jl

Lines 167 to 175 in f9cba31

	@inline ϵκx⁺⁺(i, j, k, grid, loc, κ, clock, sl, b, C) = triad_mask_x(i+1, i, j, k, k+1, grid) * κᶜᶜᶜ(i, j, k, grid, loc, κ, clock, C) * tapering_factorᶜᶜᶜ(i, j, k, grid, sl, b, C)
	@inline ϵκx⁺⁻(i, j, k, grid, loc, κ, clock, sl, b, C) = triad_mask_x(i+1, i, j, k, k, grid) * κᶜᶜᶜ(i, j, k, grid, loc, κ, clock, C) * tapering_factorᶜᶜᶜ(i, j, k, grid, sl, b, C)
	@inline ϵκx⁻⁺(i, j, k, grid, loc, κ, clock, sl, b, C) = triad_mask_x(i, i, j, k, k+1, grid) * κᶜᶜᶜ(i, j, k, grid, loc, κ, clock, C) * tapering_factorᶜᶜᶜ(i, j, k, grid, sl, b, C)
	@inline ϵκx⁻⁻(i, j, k, grid, loc, κ, clock, sl, b, C) = triad_mask_x(i, i, j, k, k, grid) * κᶜᶜᶜ(i, j, k, grid, loc, κ, clock, C) * tapering_factorᶜᶜᶜ(i, j, k, grid, sl, b, C)
	@inline ϵκy⁺⁺(i, j, k, grid, loc, κ, clock, sl, b, C) = triad_mask_y(i, j+1, j, k, k+1, grid) * κᶜᶜᶜ(i, j, k, grid, loc, κ, clock, C) * tapering_factorᶜᶜᶜ(i, j, k, grid, sl, b, C)
	@inline ϵκy⁺⁻(i, j, k, grid, loc, κ, clock, sl, b, C) = triad_mask_y(i, j+1, j, k, k, grid) * κᶜᶜᶜ(i, j, k, grid, loc, κ, clock, C) * tapering_factorᶜᶜᶜ(i, j, k, grid, sl, b, C)
	@inline ϵκy⁻⁺(i, j, k, grid, loc, κ, clock, sl, b, C) = triad_mask_y(i, j, j, k, k+1, grid) * κᶜᶜᶜ(i, j, k, grid, loc, κ, clock, C) * tapering_factorᶜᶜᶜ(i, j, k, grid, sl, b, C)
	@inline ϵκy⁻⁻(i, j, k, grid, loc, κ, clock, sl, b, C) = triad_mask_y(i, j, j, k, k, grid) * κᶜᶜᶜ(i, j, k, grid, loc, κ, clock, C) * tapering_factorᶜᶜᶜ(i, j, k, grid, sl, b, C)

@simone-silvestri @glwagner @navidcy what do you think of the tests? I don't have much experience/intuition on what should/shouldn't be in a test so your advice would be great!

[glwagner]

another missing method bug was caught during test writing:

Oceananigans.jl/src/TurbulenceClosures/abstract_scalar_diffusivity_closure.jl

Line 315 in f9cba31

@inline κᶜᶜᶜ(i, j, k, grid, loc, κ::Number, clk, fields) = κ

(funnily enough I remembered catching this bug a few weeks ago and I thought I placed a fix but I can't find where that fix was)
and here are some missing function arguments too in the diffusivity fields:

Oceananigans.jl/src/TurbulenceClosures/turbulence_closure_implementations/isopycnal_skew_symmetric_diffusivity_with_triads.jl

Lines 167 to 175 in f9cba31

	@inline ϵκx⁺⁺(i, j, k, grid, loc, κ, clock, sl, b, C) = triad_mask_x(i+1, i, j, k, k+1, grid) * κᶜᶜᶜ(i, j, k, grid, loc, κ, clock, C) * tapering_factorᶜᶜᶜ(i, j, k, grid, sl, b, C)
	@inline ϵκx⁺⁻(i, j, k, grid, loc, κ, clock, sl, b, C) = triad_mask_x(i+1, i, j, k, k, grid) * κᶜᶜᶜ(i, j, k, grid, loc, κ, clock, C) * tapering_factorᶜᶜᶜ(i, j, k, grid, sl, b, C)
	@inline ϵκx⁻⁺(i, j, k, grid, loc, κ, clock, sl, b, C) = triad_mask_x(i, i, j, k, k+1, grid) * κᶜᶜᶜ(i, j, k, grid, loc, κ, clock, C) * tapering_factorᶜᶜᶜ(i, j, k, grid, sl, b, C)
	@inline ϵκx⁻⁻(i, j, k, grid, loc, κ, clock, sl, b, C) = triad_mask_x(i, i, j, k, k, grid) * κᶜᶜᶜ(i, j, k, grid, loc, κ, clock, C) * tapering_factorᶜᶜᶜ(i, j, k, grid, sl, b, C)
	@inline ϵκy⁺⁺(i, j, k, grid, loc, κ, clock, sl, b, C) = triad_mask_y(i, j+1, j, k, k+1, grid) * κᶜᶜᶜ(i, j, k, grid, loc, κ, clock, C) * tapering_factorᶜᶜᶜ(i, j, k, grid, sl, b, C)
	@inline ϵκy⁺⁻(i, j, k, grid, loc, κ, clock, sl, b, C) = triad_mask_y(i, j+1, j, k, k, grid) * κᶜᶜᶜ(i, j, k, grid, loc, κ, clock, C) * tapering_factorᶜᶜᶜ(i, j, k, grid, sl, b, C)
	@inline ϵκy⁻⁺(i, j, k, grid, loc, κ, clock, sl, b, C) = triad_mask_y(i, j, j, k, k+1, grid) * κᶜᶜᶜ(i, j, k, grid, loc, κ, clock, C) * tapering_factorᶜᶜᶜ(i, j, k, grid, sl, b, C)
	@inline ϵκy⁻⁻(i, j, k, grid, loc, κ, clock, sl, b, C) = triad_mask_y(i, j, j, k, k, grid) * κᶜᶜᶜ(i, j, k, grid, loc, κ, clock, C) * tapering_factorᶜᶜᶜ(i, j, k, grid, sl, b, C)

@simone-silvestri @glwagner @navidcy what do you think of the tests? I don't have much experience/intuition on what should/shouldn't be in a test so your advice would be great!

I think you can just take a time-step eg

function time_step_without_error(model)
    time_step!(model, 1)
    return true
end

@test time_step_without_error(model)

or some variation thereof