Changeset 4296


Ignore:
Timestamp:
Oct 11, 2022, 10:18:49 AM (19 months ago)
Author:
Ehouarn Millour
Message:

Add "clean" initializations of coefh and coefm which are defined on klev+1 vertical levels but mostly used/computed on klev throughout the code; unitialized topmost values cause problems in outputs when in debug mode.
EM

File:
1 edited

Legend:

Unmodified
Added
Removed
  • LMDZ6/trunk/libf/phylmd/phys_state_var_mod.F90

    r4143 r4296  
    560560      ALLOCATE(coefh(klon,klev+1,nbsrf+1))
    561561      ALLOCATE(coefm(klon,klev+1,nbsrf+1))
     562      ! initialize cleanly coefh,coefm
     563      ! (most of the time in the code these are assumed to be on klev levels)
     564      coefh(:,:,:)=0
     565      coefm(:,:,:)=0
    562566      ALLOCATE(zmax0(klon), f0(klon))
    563567      ALLOCATE(sig1(klon,klev), w01(klon,klev))
Note: See TracChangeset for help on using the changeset viewer.