Changeset 519 for trunk/NEMO/OPA_SRC/TRA/traadv_ubs.F90

Timestamp:

2006-10-11T15:49:05+02:00 (18 years ago)

Author:

opalod

Message:

nemo_v1_update_74 : CT : - add comments for the UBS scheme

• syntax correction
• remove the use of eddy velocity in the calculation of poleward transport since it is already included into the normal velocity
• remove the save of the trend associated to the BBL advective velocity since it is already included into the normal velocity

File:

• trunk/NEMO/OPA_SRC/TRA/traadv_ubs.F90 (modified) (1 diff)

trunk/NEMO/OPA_SRC/TRA/traadv_ubs.F90

@@ -47 +47 @@
       !!      and add it to the general trend of passive tracer equations.
       !!
-      !! ** Method  :   ???
+      !! ** Method  :   The upstream biased third (UBS) is order scheme based 
+      !!      on an upstream-biased parabolic interpolation (Shchepetkin and McWilliams 2005)
+      !!      It is only used in the horizontal direction.
+      !!      For example the i-component of the advective fluxes are given by :
+      !!                !  e1u e3u un ( mi(Tn) - zltu(i  ) )   if un(i) >= 0
+      !!          zwx = !  or 
+      !!                !  e1u e3u un ( mi(Tn) - zltu(i+1) )   if un(i) < 0
+      !!      where zltu is the second derivative of the before temperature field:
+      !!          zltu = 1/e3t di[ e2u e3u / e1u di[Tb] ]
+      !!      This results in a dissipatively dominant (i.e. hyper-diffusive) 
+      !!      truncation error. The overall performance of the advection scheme 
+      !!      is similar to that reported in (Farrow and Stevens, 1995). 
+      !!      For stability reasons, the first term of the fluxes which corresponds
+      !!      to a second order centered scheme is evaluated using the now velocity 
+      !!      (centered in time) while the second term which is the diffusive part 
+      !!      of the scheme, is evaluated using the before velocity (forward in time). 
+      !!      Note that UBS is not positive. Do not use it on passive tracers.
+      !!      On the vertical, the advection is evaluated using a TVD scheme, as
+      !!      the UBS have been found to be too diffusive.
       !!
       !! ** Action : - update (ta,sa) with the now advective tracer trends
+      !!
+      !! Reference : Shchepetkin, A. F., J. C. McWilliams, 2005, Ocean Modelling, 9, 347-404. 
+      !!             Farrow, D.E., Stevens, D.P., 1995, J. Phys. Ocean. 25, 1731Ð1741. 
       !!----------------------------------------------------------------------
       USE oce, ONLY :   zwx => ua   ! use ua as workspace