Mixed layer depth variation with vertical resolution

[adele-morrison]

Plot maps of winter (JAS) mixed layer depth in the DSW formation regions. Does the mixed layer depth have a dependence on the top layer thickness, as discussed in this comment? Not sure if we need to look at a specific KPP boundary layer depth diagnostic or if the standard mld diagnostic would be sufficient.

[willaguiar]

This is a short analysis of mixed layer changes between models for the Weddell Sea. For OM2-MOM5 I used 1 extra year that I ran for the simulations, and outputed the hblt variable, which is suppose to be the thickness of the OSBL from KPP. For MOM6 I just used mlotst which is just the MLD from density. Notice that I've compared MLD and HBLT before for MOM5, and found very similar values. There is still a few months missing for OM2-5mtop (Sorry) but I think we can see a signal already. All densities here are $\rho_0$

Analysis below is only for the shelf (until 1km isobath) in the Weddell Sea.
On MOM5 we see that as we thicken the top cell, the OSBL on the Weddell Shelf gets thinner, and the mean density on the upper 20m of the shelf decreases ($\delta\rho [5mtop-ctrl]$ approx $0.1\ kg\ m^{-3}$).

On MOM6 the OSBL has almost no changes when increasing the thickness of the top cell. However, the density in the upper 20m of shelf still changes a little ($\delta\rho [5mtop-ctrl]$ approx $0.025\ kg\ m^{-3}$).

Why do these changes matter? It highlights that with KPP the OSBL indeed is more sensitive to the top cell thickness than on EPBL.