Setting up nuice_ref once

[abierjon]

The water ice effective variance used for radiative transfer nuice_ref seems to be set up in two different places :

• once in initracer.F (l.523 at revision r2644)

      nuice_ref=0.1   ! Effective variance nueff of the
                      ! water-ice size distribution

• once in conf_phys.F (l.709 at r2644, with the last line being a wrong comment by the way)

        write(*,*) "REF effective variance for water clouds ?"
        nuice_ref=0.45 
        call getin_p("nuice_ref",nuice_ref)
        write(*,*) "CO2 nueff Sedimentation:", nuice_ref

One of these value assignments should be discarded, as only one value is used in the simulation (but I don't know which one, it depends on the calling order for initracer and conf_phys)

[abierjon]

Some info from a svn blame on

• initracer

    38   emillour       nuice_ref=0.1   ! Effective variance nueff of the
   358      aslmd                       ! water-ice size distribution

• conf_phys

  1617  jaudouard         write(*,*) "REF effective variance for water clouds ?"
  1617  jaudouard         nuice_ref=0.45 
  2304   emillour         call getin_p("nuice_ref",nuice_ref)
  1617  jaudouard         write(*,*) "CO2 nueff Sedimentation:", nuice_ref

[abierjon]

After a quick test, the value that is kept for the run is the one from initracer (nuice_ref=0.1)

[cmathe]

resolved with commit r2659

[abierjon]

Must be reopened because r2659 only removed the 0.1 value in initracer without copying its value to nuice_ref default value in conf_phys.
If one run the GCM just after r2659, nuice_ref takes the value 0.45, while it should be 0.1 like before.

[emillour]

Indeed. Done in r2661; nuice_ref back to default value of 0.1 in conf_phys