Configuration updates to SO_wISC_10to60km_64L#510
Configuration updates to SO_wISC_10to60km_64L#510mark-petersen merged 3 commits intoMPAS-Dev:ocean/developfrom
Conversation
This is what I'm spinning up with right now. Spin-up on 4 nodes is extraorinarily slow!
TestingI tested a merge of the first commit in this PR, #506, #507, #508 and #511 together through spin-up for the following test cases:
With all 3 commits from this PR, and a test merge with #511, I tested that the |
|
In my testing of |
Kinetic Energy (log scale)No Smoothing
3 Iterations of Smoothing
6 Iterations of Smoothing
|
|
If we decide we don't want topographic smoothing globally, we would restrict it to the region near ice shelves in much the same way that the vertical coordinate is only altered from z-star in the region near ice shelves. |
Do not smooth other test cases with ice-shelf cavities by default
xylar
changed the title
mark-petersen
left a comment
mark-petersen
left a comment
There was a problem hiding this comment.
approved based on visual inspection and Xylar's testing.
Configuration updates to SO_wISC_10to60km_64L MPAS-Dev#510 Default to spinning up on 36 nodes (1296 cores) on IC. This is what I'm spinning up with right now. Spin-up on 4 nodes is extraordinarily slow! Also, switch to 6 iterations of topography smoothing. 0 and 3 iterations were producing a worrying spot of high kinetic energy near the Ross Ice Shelf, see details below. Remove an unused (and potentially confusing) template file.
3 participants
Default to spinning up on 36 nodes (1296 cores) on IC. This is what I'm spinning up with right now. Spin-up on 4 nodes is extraordinarily slow!
Also, switch to 6 iterations of topography smoothing. 0 and 3 iterations were producing a worrying spot of high kinetic energy near the Ross Ice Shelf, see details below.
Remove an unused (and potentially confusing) template file.