Changeset 2242 for trunk/LMDZ.COMMON/libf/dyn3dpar

Timestamp:

Feb 17, 2020, 5:44:12 PM (5 years ago)

Author:

jvatant

Message:

+ Update the interface with YAMMS so it now correctly handles the small values of the moments,
requiring dynamics to have a threshold quite low (set to 1D-200) and a sanity check in calmufi
corresponding to this value. Thus it removes 'most' of the unphysical radius obtained in
YAMMS. There are still some, but at least there is no more problem of model stability for the moments
and the code can run.
Still, take care the day you want to calculate opacities from the radii and not the moments.
Although, note that there are some negative values, in the output files, but theses negatives are

harmless, as they are only present in output files, dynamics reseting to epsilon after.

+ Given theses pbs with the radii, also update the optics so that it computes the opacities in
a simpler way, directly for M3, through look-up tables, M3 being a good proxy.
--JVO

File:

• trunk/LMDZ.COMMON/libf/dyn3dpar/addfi_p.F (modified) (2 diffs)

trunk/LMDZ.COMMON/libf/dyn3dpar/addfi_p.F

@@ -78 +78 @@
       REAL,PARAMETER :: qtestw = 1.0e-15
       REAL,PARAMETER :: qtestt = 1.0e-40
+      REAL,PARAMETER :: qtestt2 = 1.0D-200
 
       REAL SSUM
@@ -192 +193 @@
 c$OMP END DO
        ENDDO
+      else if (planet_type=="titan") then
+      ! Titan : needs to be able to deal with very low values of tracers
+      ! values for microphysics -> threshold at 1D-200 - JVO 20
+       DO iq = 1, nqtot
+c$OMP DO SCHEDULE(STATIC,OMP_CHUNK)
+         DO k = 1,llm
+            DO j = ijb,ije
+               pq(j,k,iq)= pq(j,k,iq) + pdqfi(j,k,iq) * pdt
+               pq(j,k,iq)= AMAX1( pq(j,k,iq), qtestt2 )
+            ENDDO
+         ENDDO
+c$OMP END DO
+       ENDDO
       else
       ! general case, treat all tracers equally)