Surface water vapor content variables: Hair and Qair

[ganixpan]

Hi all:

While making some changes to a generic varinfo.dat file to use WRF output as atmospheric forcing in the bulk fluxes calculations, I've got confused with the name of the surface water vapor content variable.
Looking in a generic varinfo.dat file one can find that the 'Qair' variable is related to that content of vapor (Relative Humidity):

Code: Select all

'Qair'                                             ! Input
  'surface air relative humidity'
  'percentage'                                     ! [percentage]
  'Qair, scalar, series'
  'qair_time'
  'idQair'
  'r2dvar'
  0.01d0                                           ! 1/100

However, when looking in bulk_flux.F the name related to the water vapor content has changed to 'Hair', although it seems to be Relative Humidity yet:
in subroutine definition

Code: Select all

      SUBROUTINE bulk_flux_tile (ng, Istr, Iend, Jstr, Jend,            &
     &                           LBi, UBi, LBj, UBj,                    &
     &                           nrhs,                                  &
# ifdef MASKING
     &                           rmask, umask, vmask,                   &
# endif
     &                           alpha, beta, rho, t,                   &
     &                           Hair, Pair, Tair, Uwind, Vwind, ....

and in input field definitions:

Code: Select all

!
!  Input bulk parameterization fields.
!
          ...
          rhoSea(i)=rho(i,j,N(ng))+1000.0_r8
          RH=Hair(i,j)
          SRad(i,j)=srflx(i,j)*Hscale
          ...

A bit later in bulk_flux.F 'Qair' is there again and now it is a Saturation Specific Humidity!!!

Code: Select all

!    note that Qair(i) is the saturation specific humidity at Tair
!                 Q(i) is the actual specific humidity
!              Qsea(i) is the saturation specific humidity at Tsea
!
!          Saturation vapor pressure in mb is first computed and then 
!          converted to specific humidity in kg/kg

I suppose that 'Qair' name in the varinfo.dat file is internally changed to 'Hair' and later re-used for another variable, but anyway, this could be a problem with variable name 'Qair' and its units so...

Thank you!!

[kate]

Yes, this is terribly confusing. If you look inside bulk_flux.F, you see that it is unsure of the units of the incoming humidity. Is it relative or specific humidity? I know I'm using the CORE forcing files which provide specific humidity in MKS units, so there's no need for the conversion from relative to specific humidity. There should be a switch for that.