b.e30_alpha08b.B1850C_LTso.ne30_t232_wgx3.314

[cecilehannay]

Purpose:

Adding the new changes from 311 to 308 and revert r_snw to 1.

---

Description:
Same as 308 (#290) except:

• user_nl_cice

r_snw = 1.25 -> 1.0

• use ian ice runoff's file (250km ice runoff radius + mask to avoid pile-up of ice)

./xmlchange ROF2OCN_ICE_RMAPNAME=/glade/u/home/igrooms/runoff_mapping/map_r05_to_tx2_3_nnsm_e250r250_modified_260212.nc
./xmlchange ROF2OCN_LIQ_RMAPNAME=/glade/work/gmarques/cesm/tx2_3/runoff_mapping/map_r05_to_tx2_3_nnsm_e100r100_250327.nc

---

Case directory:

• Locally (if still available):
  /glade/campaign/cesm/cesmdata/cseg/runs/cesm2_0/b.e30_alpha08b.B1850C_LTso.ne30_t232_wgx3.314

• On github:
  https://github.com/NCAR/cesm_dev/tree/b.e30_alpha08b.B1850C_LTso.ne30_t232_wgx3.314

---

Sandbox:

• Locally (if still available): /glade/work/hannay/cesm_tags/cesm3_0_alpha08b

• On github: https://github.com/ESCOMP/cesm/tree/cesm3_0_alpha08b

---

Diagnostics:

• Diags (if available):
  https://webext.cgd.ucar.edu/BLT1850/b.e30_alpha08b.B1850C_LTso.ne30_t232_wgx3.314/

---

Output:

• Raw output on derecho (if still available): /glade/derecho/scratch/hannay/archive/b.e30_alpha08b.B1850C_LTso.ne30_t232_wgx3.314/
• CAM Climos: /glade/campaign/cgd/amp/amwg/climo/b.e30_alpha08b.B1850C_LTso.ne30_t232_wgx3.314

---

Contacts:
@cecilehannay @dlawrenncar @adamrher @PeterHjortLauritzen @JulioTBacmeister @swrneale @slevis-lmwg @wwieder @olyson @gustavo-marques @iangrooms @fobryan3 @klindsay28 @dabail10 @duvivier @tilmes @mbramberger @gunterl

---

Extra details:

[dabail10]

Sea ice diagnostics. Maybe it is related to the runoff? This volume in the Arctic is lower than 302. We could probably do r_snw = 1.25 here if the TOA needs it.

https://webext.cgd.ucar.edu/BLT1850/b.e30_alpha08b.B1850C_LTso.ne30_t232_wgx3.314/ice/html/infrastructure/index.html

[dabail10]

Actually, I do like the Arctic volume here better than 302.

[duvivier]

@iangrooms
Adding this here too: I spoke to Dave about this hand adjusted re-mapping file for runoff. Seems reasonable to me to prevent build up in embayments. Can you provide a map highlighting the points that you re-routed?

Also, can you clarify what happens to the re-routed runoff. For example, does Ross Island runoff get re-routed just to nearby points but spread some, or does it get mapped over the whole southern ocean?

[iangrooms]

Added some info to the slides for today, but I will also add some here.

There are two grids for runoff: a source grid (evidently 1/2 degree lat/lon) and a target grid (ocean tx2_3 v2). Each cell in the source grid spreads its runoff to a subset of cells in the target grid. Some cells on the target grid have excessive ice buildup. I have, in an ad hoc process, identified 31 'bad' cells in the target grid, and I modified the runoff map so that there is no runoff into these bad cells.

Consider a single bad cell in the target grid. Lots of cells in the source grid all put runoff into that bad target cell. Consider just one cell from the source grid that puts runoff into a bad target cell. The runoff map spreads from the one source grid cell into many target grid cells, including the bad one. I modify the map so that there is no runoff from that one source grid cell into the bad target grid cell. But I want to maintain conservation, so I scale the runoff from that source cell into the other target cells in a way that ensures conservation. I do not put runoff from that source cell into any new cells in the target grid.

The process of modifying the runoff map takes two things: an existing runoff map and a list of bad points in the target grid, and it returns a modified runoff map. There's a Jupyter notebook and a Matlab script that together do the modification, so it's reproducible. It operates via a nested loop: For every bad target cell it finds all the source cells that put runoff into that cell; then for every one of those source cells it zeros out the runoff to the bad cell and inflates the runoff to the other cells.

[duvivier]

For the record, I am including some figures from Ian here of the mapping for the 31 points. The runoff from a "bad" cell is conservatively rerouted to other cells within the "bad" cell's radius zone.

[cecilehannay]

ADF 314 vs 308:
https://webext.cgd.ucar.edu/BLT1850/b.e30_alpha08b.B1850C_LTso.ne30_t232_wgx3.314/atm/