Order of calls in icepack_therm_vertical.F90

[TillRasmussen]

Should the neutral_drag_coeffs subroutine be called before the frzmlt_bottom_lateral subroutine within the icepack_therm_vertical module? neutral_drag_coeffs update the ocean drag which is used in the frzmlt_bottom_lateral, however the frzmlt_bottom_lateral routine is called first which means that Cdn_ocn from the previous time step is used-

[eclare108213]

Wow. I'll need to dig a bit to understand why this was done, or if it was always this way. Did you try running it both ways, to see how much of a difference it makes?

[dabail10]

This subroutine is only called when formdrag is on. This is a relatively new formulation and I don't believe it has been blessed. However, it probably should be called before frzmlt_bottom_lateral when formdrag is .true.

[TillRasmussen]

@eclare108213 I did not try to rerun swapping the order, however I think that the change will be small. The current setup will utilize the ocean drag from the previous time step or initialization and only when formdrag is on. I think that the biggest difference is when the model is cold started/restarted. @dabail10 time is a relative measure :) The reason I got to look at this was based on this study https://agupubs.onlinelibrary.wiley.com/doi/pdf/10.1029/2020JC016977 which claim that the ocean drag in summer was over estimated in summer due to floe size drag. In addition when turned on the DMI model has a tendency of opening up a polynya north of Greenland.

[TillRasmussen]

The change passed the qc test. I will make a pull request

[eclare108213]

merged in PR #381