qmin for chemistry/aerosol species read in from snapshot are incorrectly initialized in initialize_constituents

[jimmielin]

This code in current CAM, src/chemistry/mozart/chemistry.F90

       qmin = 1.e-36_r8
...
       if ( lng_name(1:5) .eq. 'num_a' ) then ! aerosol number density
          qmin = 1.e-5_r8
       else if ( m == o3_ndx ) then
          qmin = 1.e-12_r8
       else if ( m == ch4_ndx ) then
          qmin = 1.e-12_r8
       else if ( m == n2o_ndx ) then
          qmin = 1.e-15_r8
       else if( m == cfc11_ndx .or. m == cfc12_ndx ) then
          qmin = 1.e-20_r8
       else if( m == o2_1s_ndx .or. m == o2_1d_ndx ) then

is not handled in initialize_constituents scheme, resulting in answer changes due to a different qmin. Currently, initialize_constituents initializes qmin for all species as 0.0_kind_phys

This will have to be resolved in two places:

1. for snapshot runs, initialize_constituents needs to replicate the above logic, and
2. when chemistry constituents are actually initialized for a real model run (not in this issue)

I will work on (1) which is the scope of the issue here, as part of the UW PBL scheme work

cc @peverwhee