Vertically stretched Cartesian grid

[ali-ramadhan]

This PR adds a vertically stretched grid.

Resolves #47
Works towards #471
Resolves #531
Resolves #544

[codecov]

Codecov Report

Merging #543 into master will increase coverage by 5.14%.
The diff coverage is 63.05%.

@@            Coverage Diff             @@
##           master     #543      +/-   ##
==========================================
+ Coverage   68.64%   73.79%   +5.14%     
==========================================
  Files          70       71       +1     
  Lines        1990     2049      +59     
==========================================
+ Hits         1366     1512     +146     
+ Misses        624      537      -87

Impacted Files	Coverage Δ
src/Operators/derivative_operators.jl	90.62% <ø> (+32%)	⬆️
src/Operators/divergence_operators.jl	100% <ø> (ø)	⬆️
src/Operators/tracer_advection_operators.jl	100% <ø> (ø)	⬆️
...urbulenceClosures/viscous_dissipation_operators.jl	100% <ø> (ø)	⬆️
src/Solvers/Solvers.jl	100% <ø> (ø)	⬆️
src/Operators/laplacian_operators.jl	100% <ø> (+85.71%)	⬆️
src/OutputWriters/jld2_output_writer.jl	84.61% <ø> (ø)	⬆️
src/Operators/momentum_advection_operators.jl	100% <ø> (ø)	⬆️
src/Diagnostics/cfl.jl	66.66% <ø> (ø)	⬆️
src/Oceananigans.jl	75% <ø> (ø)	⬆️
... and 40 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 9fbbc93...8e51fdc. Read the comment docs.

[ali-ramadhan]

If we're okay with merging a vertically stretched grid abstraction by itself with some tests that ensure the vertical levels are correctly computed then I think this PR is ready to be reviewed.

Pressure solver and full integration will come in a future PR.