Changeset 3333 for trunk/LMDZ.MARS/libf/phymars/vdif_cd_mod.F90

Timestamp:

May 17, 2024, 11:30:17 AM (6 months ago)

Author:

llange

Message:

Mars PCM
Fixing a bug in vdif_cd: a "residual", used as criterion to enter an iterative loop, was wrongly initialized. Hence, for some points, the algorithm does not go into the loop, and a wrong value of Cd, Ch was computed.
Also some cleaning/small fixing with save variables with OMP.

LL

File:

• trunk/LMDZ.MARS/libf/phymars/vdif_cd_mod.F90 (modified) (3 diffs)

trunk/LMDZ.MARS/libf/phymars/vdif_cd_mod.F90

@@ -157 +157 @@
       pcdh(:,:) = 0.
       z0t(:,:) = 0.
+      fm(:,:) = 0.
+      fh(:,:) = 0.
       f_ri_cd_min = 0.01
 ! this formulation assumes alphah=1., implying betah=betam
@@ -203 +205 @@
             DO ig=1,ngrid
                ite = 0.
-               residual = abs(pz0tcomp(ig)-pz0t) 
+               residual = 100*tol_iter*pz0(ig) 
                z1z0=pz(ig,1)/pz0(ig)
                cdn(ig)=karman/log(z1z0)
                cdn(ig)=cdn(ig)*cdn(ig)
-            
                DO WHILE((residual .gt. tol_iter*pz0(ig)) .and.  (ite .lt. itemax)) 
 ! Computations of z0T; iterated until z0T converges  
@@ -225 +226 @@
                       rib_dry(ig,islope) = ric_colaitis
                    ENDIF
-        
 ! Compute the stability functions fm; fh depending on the stability of the surface layer