Changes between Version 42 and Version 43 of ticket/1331_UKMO2_ice_shelves

Timestamp:

2014-11-28T17:16:42+01:00 (9 years ago)

Author:

mathiot

Comment:

--

ticket/1331_UKMO2_ice_shelves

@@ -60 +60 @@
      - configuration ISOMIP   vvl after 133 days with  constant  flux: conservation of t,s,vol is much much better (<~1e-10).
      - comments: I have no idea why it is better if specified flux and constant flux. I give up for the moment on it.
- * Restartability tested with ISOMIP: Restartability works only with key_ldfslp activated even if tra/dynldf_iso set to false !!!! if not key_ldfslp, see #1359: Restartability issue without key_ldfslp ticket. Processor configurations tested: 1x8; 8x1; 1x1; 8x4; 4x8; in ISOMIP configuration (12 steps).
+ * Restartability tested with ISOMIP: Restartability works only with key_ldfslp activated even if tra/dynldf_iso set to false !!!! if not key_ldfslp, see #1359: Restartability issue without key_ldfslp ticket.
  * option compatibility with ice shelf cavity:
      - STOP if sbccpl interface used (could be a mask issue somewhere in cpl module, not tested)
@@ -84 +84 @@
  * Which routine(s) are causing the difference?
    - domzgr.F90: For ice shelves, I need to define the the e3t/u/v/w=zt/u/v/w(k+1) - zt/u/v/w(k). If not, dep3w(i,j,k) (could be) .NE. dep3w(i,j+1,k) in the full wet cell beneath the ice shelf. 
-   - zpshde.F90
-   - dynhpg.F90
+   - zpshde.F90 (2 cell water columns issue with the reference file)
+   - dynhpg.F90 (only if you used hpg_sco.)
 This has been only tested in GYRE configuration with bump at the bottom and ln_zps activated.
  * Why the changes are not protected by a logical switch or new section-version?