Update global ocean init mode for meshes with ice-shelf cavities #90
Conversation
|
This merge works in tandem with MPAS-Dev/compass#813. |
| call ocn_equation_of_state_density(statePool, meshPool, tracersSurfaceValue, & | ||
| nCells, 0, 'relative', density, iErr, & | ||
| timeLevelIn=1) |
There was a problem hiding this comment.
Am I right in thinking that this density calculation doesn't use land ice pressure (atm pressure everywhere)? So as far as this calculation is concerned, ice shelf cavities don't exist yet (ssh = 0)?
There was a problem hiding this comment.
It uses a reference pressure profile that doesn't vary at all with space. It doesn't even use atmospheric pressure. (It's really kind of an embarrassment.)
| landIcePressure(:) = 0.0_RKIND | ||
| modifyLandIcePressureMask(:) = 0 |
There was a problem hiding this comment.
It would be nice if this were called sshAdjustmentMask since we're not always modifying landIcePressure, but it's up to you if you want to bother.
There was a problem hiding this comment.
Sure, this will require changes outside of global_ocean but I'm okay with that.
There was a problem hiding this comment.
I have done this both here and in MPAS-Dev/compass#813
xylar
force-pushed
the
ocn/init-land-ice-pressure-from-topo
branch
from
cea5af4 to
d9cb98e
Compare
|
@cbegeman, can you let me know if you think this is ready for E3SM-Project/E3SM yet? |
|
@xylar Would you like me to do any testing? If so, are there machines/compilers you haven't tried yet? If not, then feel free to move to E3SM. Thanks for addressing my comments! |
|
I've only run with Intel on Chrysalis so it wouldn't hurt to run a test elsewhere. I wouldn't try anything higher rest than IcoswISC and IcoswISC240 might even be sufficient. But I'd certainly appreciate some testing if you're up for it. |
We do not use this mode anymore and it is too difficult to maintain.
This merge uses the `landIceDraftObserved`, `landIcePressureObserved` and `landIceFloatingFracObserved` to initialize the corresponding MPAS-Ocean variables. Previously, the `landIcePressure` was being computed in a subroutine and the `landIceFloatingFraction` was required to be identical to `landIceFraction`. This merge also removes the unused variable `landIceGroundedFracObserved`
Instead, it will be read in from the topography file directly, meaning it can be updated (e.g. from ssh adjustment).
xylar
force-pushed
the
ocn/init-land-ice-pressure-from-topo
branch
from
d9cb98e to
114c9a3
Compare
|
Closed in favor of E3SM-Project#6375 |
This merge updates MPAS-Ocean init mode for
global_oceanin several ways that are critical for initialization that is compatible with planned coupling to MALI.The fields
landIcePressure,landIceDraftandsshare now read from a topography file. This means thatlandIcePressureis no longer computed based on the density at the ocean surface in init mode. Similarly,sshis not required to be the same aslandIceDraft.Support is removed for reading land-ice topography fields from a dataset on a lat-lon grid. We no longer use this capability and maintaining it is not worth the effort.
The land-ice topography data set is now required to have
landIceFloatingFracObservedas a field, and this is used to setlandIceFloatingFraction. Previously,landIceFracObservedwas being used to compute bothlandIceFractionandlandIceFloatingFraction(assuming no grounded ice in the domain).