Changes between Version 11 and Version 12 of 2009WP/2009Stream3/VVL

Timestamp:

2009-06-21T11:42:40+02:00 (15 years ago)

Author:

gm

Comment:

--

2009WP/2009Stream3/VVL

@@ -80 +80 @@
             solve dU/dt = ...  and not d( e3 U)/dt = ... ). [[BR]]
             Change to be done in both dynnxt AND dynspg (_ts and flt)
+
             '''In dynnxt.F90'''   perform the thickness weighted time leap-frog and  Asselin filter only in ln_dynadv_vec=FALSE and lk_vvl=TRUE., in other word unweighted operation if  n_dynadv_vec=T and NOT lk_vvl. So replace the 2 two [[BR]]
 
@@ -99 +100 @@
 }}}
 
-    In addition : (1) the masking of ua, va can be moved inside the key_bdy case ; (2) all the comment have been updated. The resulting subroutine can be found in attachment ().
-          
+    In addition : (1) the masking of ua, va can be moved inside the key_bdy case ; (2) all the comment have been updated. The resulting subroutine can be found in attachment (https://forge.ipsl.jussieu.fr/nemo/attachment/wiki/2009WP/2009Stream3/VVL/dynnxt.F90).
+
+   '''In dynspg_ts.F90'''   the whole module have to be modified to solve a velocity equation, not a thickness weighted velocity equation. Then in the specific case of ln_dynadv_vec=FALSE and lk_vvl=TRUE, an term associated with the time derivative of the ssh must be added to the equation solved. [[BR]]
+
  - ensure that BDY and OBC are working with the new time-stepping strategy
         in particular obc_dyn_bt  seems to modify the sshn field. This should be wrong with the new time strategy