Apply dss to slab model Y+dY

[LenkaNovak]

Address issues in #380 in all slab models.

Components

• implement dss! hooks in model-specific init files for ocean and sea sea ice slab models
• note difference in surface fluxes - expecting a smoother field
• ensure Buildkite drivers produce physically sensible results, note performance changes

Below we use an example of the idealized helem=4 aquaplanet

• 1. surface flux differences - fairly subtle (compared to the land ones in Apply dss to land Y+dY  #380 )

energy fluxes without / with DSS (W/m2)

difference

• 2. performance - small increase in allocs and walltime (as expected from additional communication)