incorrect direct call to setFilters

[bandre-ucar]

Summary of Issue:

Email from Bill Sacks:

I have a vague recollection that I have mentioned this already, so apologies if I'm repeating myself: I noticed that there is a direct call to setFilters in the driver for the ED code. If it's really needed (which I kind of doubt at this point, but I'm not at all sure), then it should be replaced by a call to reweightWrapup – which does some other stuff additional to setting the filters.

Sorry: that should be reweight_wrapup.

Also, I think there's a threading bug in that call to setFilters, in that there should be bounds specified on the array. (Though I'm about to change what that call looks like in an upcoming trunk tag....)

NOTE(bja): a threading issue could account for #14

Expected behavior and actual behavior:

see above

Steps to reproduce the problem (should include create_newcase or create_test command along with any user_nl or xml changes):

n/a

What is the changeset ID of the code, and the machine you are using:

any ED version based on >= clm4_5_x_r120

have you modified the code? If so, it must be committed and available for testing:

n/a

Screen output or output files showing the error message and context:

n/a

[rgknox]

The setFilters subroutine thinks that the array that is passed to it is starting with an index of bounds_clup%begg. So unless the array that is passed in, has been prepared with that expectation some-time before the call, then it does look as though it is not starting at bounds_clump%begg.

This seems like a simple enough test to pass glc2lnd_inst%icemask_grc(bounds_clump%begg:bounds_clump%endg) instead of glc2lnd_inst%icemask_grc.

      if ( use_ed  .and. is_beg_curr_day() ) then ! run ED at the start of each day

          if ( masterproc ) then
             write(iulog,*)  'clm: calling ED model ', get_nstep()
          end if

          call ed_driver( bounds_clump,                               &
                ed_allsites_inst(bounds_clump%begg:bounds_clump%endg), &
                ed_clm_inst, ed_phenology_inst,                        &
                atm2lnd_inst, soilstate_inst, temperature_inst,        &
                waterstate_inst, canopystate_inst)

          call setFilters( bounds_clump, glc2lnd_inst%icemask_grc )

       end if ! use_ed branch

 !------------------------------------------------------------------------
  subroutine setFilters(bounds, icemask_grc)
    !
    ! !DESCRIPTION:
    ! Set CLM filters.
    use decompMod , only : BOUNDS_LEVEL_CLUMP
    !
    ! !ARGUMENTS:
    type(bounds_type) , intent(in) :: bounds
    real(r8)          , intent(in) :: icemask_grc( bounds%begg: ) ! ice sheet grid coverage mask [gridcell]
    !------------------------------------------------------------------------

[rgknox]

I changed the argument in the SetFilters call to include the bounds_clump:begg. It had no effect on the success rate of our regression tests. I think we have a bunch of minor issues out there with simple fixes.
How do you guys feel about staging them in a series of commits and submitting them as 1 PR?

#29 #23
#44
#36
#4

[rgknox]

even though tests did not seem to show an effect, this needs to be de-wronged anyway. I will get the PR submitted with the other minor fixes c_area and bb_slope.