Internal wave drag implementation #607
Conversation
sbrus89
force-pushed
the
ocean/coastal
branch
2 times, most recently
from
025c036 to
acea49c
Compare
|
This is going to be canceled following advice from @nairita87, a new PR will be submitted. |
nairita87
force-pushed
the
topographic_wave_drag
branch
from
340cf8f to
51e4825
Compare
proteanplanet
added
tides
nairita87
changed the title
mark-petersen
force-pushed
the
topographic_wave_drag
branch
from
4c5e30e to
8084b70
Compare
nairita87
force-pushed
the
topographic_wave_drag
branch
from
8084b70 to
db86687
Compare
|
Rebased to head of ocean/develop. Passes nightly test suite. |
|
@nairita87: I think this is ready to merge? We just need @mark-petersen to review? |
|
The PR has been verified against previous studies (Pringle et al 2021) and also with TPXO8 data. The following plot can be compared with Pringle et al 2021 (Figs 5 and 6) Below is the plot showing code agreement before and after rebasing |
… compilation bug)
The wave drag is read from the HYCOM file in subroutine src/core_ocean/shared/mpas_ocn_vel_forcing_topographic_drag.F
mark-petersen
force-pushed
the
topographic_wave_drag
branch
3 times, most recently
from
3edb6bd to
0b04b8a
Compare
|
Verified the rebased code with the latest commit #a645c40. Matches well with previous versions |
mark-petersen
force-pushed
the
topographic_wave_drag
branch
from
be0bcf3 to
0084cdc
Compare
mark-petersen
force-pushed
the
topographic_wave_drag
branch
from
0084cdc to
a2242cd
Compare
| ! du/dt = ... (1.0/topographic_wave_drag_rinv) * u | ||
| ! topographic_wave_drag_rinv is the e-folding time which is read from a file | ||
| tend(k,iEdge) = tend(k,iEdge) - topographicWaveDragCoeff & | ||
| /(0.5_RKIND*(topographic_wave_drag_rinv(cell1)+topographic_wave_drag_rinv(cell2)))*normalVelocity(k,iEdge) |
There was a problem hiding this comment.
@nairita87 on these lines we have a divide and average from cell to edge every time step. It makes a lot more sense compute rinv in your original python interpolation to edges, and save a field which is 1/rinv at the edges. Call it topographic_wave_drag. Then these lines would simply be
131 tend(k,iEdge) = tend(k,iEdge) - topographicWaveDragCoeff * &
topographic_wave_drag(iEdge)*normalVelocity(k,iEdge)
ab35511 to
e51fae3
Compare
mark-petersen
force-pushed
the
topographic_wave_drag
branch
2 times, most recently
from
6a49d30 to
52dfe32
Compare
There was a problem hiding this comment.
@philipwjones the boundary edges between water and land (i.e. no cells) have maxLevelEdgeTop(iEdge)=0, so this if statement was needed within the edge loop to prevent accessing k=0. Is there a more performant way to do this?
There was a problem hiding this comment.
Probably not worth worrying about for this small loop and the alternatives are probably less robust, so I think ok to keep here.
mark-petersen
force-pushed
the
topographic_wave_drag
branch
from
d7cce4a to
554a53e
Compare
Internal wave drag implementation #607
|
Code verification against old PR, |
Merge PR #607 'nairita87/topographic_wave_drag' into ocean/develop
…xt (PR #4255) Update mpas-source: ocean drag parameterizations Brings in a new mpas-source submodule with changes only to the ocean core, along with correspodning updates to the mpas-ocean bld scripts. It adds wave drag parameterization and bottom roughness, which are by default off and should not impact E3SM. See MPAS-Dev/MPAS-Model#607 [NML] [BFB]
mark-petersen
left a comment
mark-petersen
left a comment
There was a problem hiding this comment.
Looks good. @nairita87 thanks for your work.
Update mpas-source: ocean drag parameterizations Brings in a new mpas-source submodule with changes only to the ocean core, along with correspodning updates to the mpas-ocean bld scripts. It adds wave drag parameterization and bottom roughness, which are by default off and should not impact E3SM. See MPAS-Dev/MPAS-Model#607 [NML] [BFB]
Merge PR MPAS-Dev#607 'nairita87/topographic_wave_drag' into ocean/develop
5 participants
Topographic wave drag term added to the momentum equation.
When
config_use_topographic_wave_drag = .true.we add a drag termdu/dt= -(1/rinv)*u, where 'rinv' is the e-folding time-scale. The e-folding time-scale is read from a map (in this case from the HYCOM data)When
config_use_implicit_bottom_roughness = .trueuse the bottom drag formulationmax(cbmin, (vonk/log(0.5*depth(i,j)/z0))**2 ))from NCOM.