MHS Obs Validation

[ADCollard]

Need to validate observations from MHS and evaluate QC procedures in the MHS YAML file, as a MHS YAML file already exists in the GDASApp.

This issue will be used to track progress on validation of radiances in UFO for GDAS in four stages:

• [x] HofX computation using GSI produced GeoVaLs
• [x] QC count checks (and YAML changes) using GSI produced GeoVaLs
• [x] Obs errors using GSI produced GeoVaLs
• [ ] HofX computation in FV3-JEDI

[ADCollard]

This is the missing ifail_fact1 QC check:

  if(sea .or. ice .or. snow)then
     dsi=9.0_r_kind
     if(tb_obs(2) < h300)then
        dsi=0.13_r_kind*(tbc(1)-33.58_r_kind*tbc(2)/(h300-tb_obs(2)))
!       QC3 in statsrad
        if(luse .and. dsi >= one)aivals(10,is) = aivals(10,is) + one
     end if
!    si=42.72_r_kind+0.85_r_kind*tbc(1)-tbc(2)
  else
     dsi=0.85_r_kind*tbc(1)-tbc(2)
!    si=42.72_r_kind+0.85_r_kind*tb_obs(1)-tb_obs(2)
!    QC4 in statsrad
     if(luse .and. dsi >= one)aivals(11,is) = aivals(11,is) + one
  end if
  dsi=max(zero,dsi)
  fact1=((tbc(1)-7.5_r_kind*dsi)/r10)**2+(dsi)**2

  if(fact1 > one)then
     vfact=zero
!    QC1 in statsrad
     if(luse)aivals(8,is) = aivals(8,is) + one
     do i=1,nchanl
        if(id_qc(i) == igood_qc)id_qc(i)=ifail_fact1_qc
     end do
  else
     if (mhs) then  ! wv sounding channels
        do i=3,nchanl
           if (abs(tbc(i)) >= two) then
              varinv(i) = zero
              if(id_qc(i) == igood_qc)id_qc(i)=ifail_gross_routine_qc
           end if
        end do
     end if
     efact = (one-fact1*fact1)*efact
     vfact = (one-fact1*fact1)*vfact

I don't think we can do this in YAML. So it will require new code and ctests!

Or we can see what the exisiting Bennartz Scattering Test does.

[azadeh-gh]

[azadeh-gh]

We did not inflate error using mhs scattering index (similar to GSI) , but used Bennartz Scattering QC check in JEDI YAML file. So the obs errors are different between JEDI and GSI.

[azadeh-gh]

mhs_metob_b GSI vs JEDI HofX

[azadeh-gh]

QC count check

[azadeh-gh]

[azadeh-gh]

OmF

[azadeh-gh]

Obs error check

[azadeh-gh]

mhs_n19 JEDI vs GSI HofX were identical

[azadeh-gh]

QC checks

[azadeh-gh]

Obs Error

[azadeh-gh]

OmF difference