about QCORRECTION

[lmkli]

Hello,

Is dqdt in set_vbc.F same as dQdSST in forcing file?

stflx(i,j,itemp)=stflx(i,j,itemp)+ &
& dqdt(i,j)*(t(i,j,N(ng),nrhs,itemp)-sst(i,j))

If so, the unit of stflx is "degC m/s", while the unit of dQdSST is "Watts meter-2 Celsius-1".
It seems the units of both sides of the equation are not matching.

If not, what's the relationship between them?

Thanks for any reply.

[meiw]

I think they are the same. When you set the value of dqdt in "ana_sst.h", you need to divide it by (rho0*Cp), and then the unit of dqdt is m/s.

[chimera.u]

In my ana_sst.h

!-----------------------------------------------------------------------
! Set sea surface temperature (Celsius) and heat flux sensitivity to
! SST (Watts/m2).
!-----------------------------------------------------------------------
!
DO j=JstrR,JendR
DO i=IstrR,IendR
sst(i,j)=???
dqdt(i,j)=???
END DO
END DO
ana_sst.h: no values provided for sst and dqdt.

and

! expressions. The forcing dQdSTT is usually computed in units of !
! (Watts/m2/degC). It needs to be scaled to (m/s) by dividing by !
! rho0*Cp.

so easy to miss it..., what's the consequence if just keep it the way as dqdt(i,j)=???.... during the compiling there is no warning

[lmkli]

I think they are the same. When you set the value of dqdt in "ana_sst.h", you need to divide it by (rho0*Cp), and then the unit of dqdt is m/s.

Do I need to divide dQdSST by (rho0*Cp) when I am preparing the forcing file?

In the varinfo.dat:

'dQdSST' ! Input
'surface net heat flux sensitivity to SST'
'watt meter-2 Celsius-1' ! Input: [Watt/m2]
'dQdSST, scalar, series' ! [Celsius m/s]
'sst_time'
'iddQdT'
'r2dvar'
1.0d0

I think it suggests that I don't need to divide it by (rho0*Cp) when I prepare dQdSST

[meiw]

In my ana_sst.h

!-----------------------------------------------------------------------
! Set sea surface temperature (Celsius) and heat flux sensitivity to
! SST (Watts/m2).
!-----------------------------------------------------------------------
!
DO j=JstrR,JendR
DO i=IstrR,IendR
sst(i,j)=???
dqdt(i,j)=???
END DO
END DO
ana_sst.h: no values provided for sst and dqdt.

and

! expressions. The forcing dQdSTT is usually computed in units of !
! (Watts/m2/degC). It needs to be scaled to (m/s) by dividing by !
! rho0*Cp.

so easy to miss it..., what's the consequence if just keep it the way as dqdt(i,j)=???.... during the compiling there is no warning

I think this depends on whether "QCORRECTION" is defined in your header file or not.

[meiw]

I think they are the same. When you set the value of dqdt in "ana_sst.h", you need to divide it by (rho0*Cp), and then the unit of dqdt is m/s.

Do I need to divide dQdSST by (rho0*Cp) when I am preparing the forcing file?

In the varinfo.dat:

'dQdSST' ! Input
'surface net heat flux sensitivity to SST'
'watt meter-2 Celsius-1' ! Input: [Watt/m2]
'dQdSST, scalar, series' ! [Celsius m/s]
'sst_time'
'iddQdT'
'r2dvar'
1.0d0

I think it suggests that I don't need to divide it by (rho0*Cp) when I prepare dQdSST

I think you still need to divide it by (rho0*Cp) in your forcing file. Actually, it is easy to check whether this is right or not since (rho0*Cp) is a very big number. You can run and compare two experiments: one with dqdt divided by (rho0*Cp) and the other without.