Enforce a range of layer interfaces to constant minimum pressure levels for BLOM hybrid vertical coordinate #432
Conversation
…layer interfaces are enforced to constant pressure levels.
|
I'm happy to test this for the NOINYOC compsets. I would suggest to do this after this has been merged (since the old behavior can be easily reproduced by namelist changes) |
There was a problem hiding this comment.
@matsbn - Thanks! Looks fine to me.
One question: Is the code block starting from line 625 in mod_ale_regrid_remap.F90 : do while (kt <= kdmx(i)) sometimes skipped (i.e. can we have kdmx(i) < kt in some locations)?
It may happen that kdmx(i) < kt. Then all destination interfaces for the regridding, p_dst(k,i), with k >= kt will be set to the pressure of the deepest interface (at the ocean floor). With k_range_plevel = 1, this would mean that the whole water column is placed in the top layer (k = 1), while k_range_plevel > 1 will ensure some resolution near the surface. |
|
Related to this, is there a minimum layer thickness that is maintained? |
Yes, for interior layers (not layers that can have zero thickness at the ocean floor), the minimum thickness is controlled by the namelist variable dpmin_interior. Currently this is set to 0.1 m (converted to a pressure thickness in the code). |
Currently for BLOM hybrid vertical coordinate, a water column lighter that the reference potential density of the k = 2 interface would occupy only the first model layer. With this PR, a range of layer interfaces with indices [1, k_range_plevel] will be enforced to the prescribed constant minimum pressure levels, always ensuring some vertical resolution near the surface. Setting the namelist variable k_range_plevel = 1, reproduces the old behaviour. The default value of k_range_plevel has been set to 4 in NorESM configurations, which changes results for vcoord = 'cntiso_hybrid'.
@TomasTorsvik and @JorgSchwinger, it would be interesting to see if this PR would mitigate some of the issues reported in #390.