Evidence that all of the right pieces are showing up. The blanked cells in blue, the drag cells in red, the ABL flow above the waves, and the wave velocity profile as the target velocity within the blanked cells (arrows).

Working for the waves2amr pathway, i.e. using wave modes from HOS-Ocean. Getting the right velocity inside the waves requires the bug fix #1372 (already included).

Will be meeting with Harish to hash out the question marks on the surface + turbulence modeling sides.

After discussing with Harish, we concluded that this PR is mostly there. The main thing that remains is adapting the wall-model forcing term (applied to cells flagged as "drag" cells). We may consult Manuel Ayala's work for this.

Other possible concerns are the magnitude of the Cd term (applied to cells flagged as "blanked" cells) - too low may mean that the velocity within the waves doesn't change fast enough, and too high may mean stability issues - and overlap of waves and sponge forcing near outlets - may need to smooth waves at inflow or outflow. But these concerns have to do with robustness rather than accuracy.

Other design choices: for now, we are assuming the ocean surface has a uniform temperature in a single simulation. If the simulations get bigger, we may need to change that. We are also assuming a uniform roughness over the waves (z0). This parameter represents the underresolved waves, not the resolved ones that are modeled as terrain.