units of omega

[jpringle]

Dear all --

I, and some Europeans doing particle tracking with ROMS output, are confused by what the units of omega are supposed to be in the output. From equation 12 of https://www.myroms.org/wiki/Terrain-Fol ... sformation , the units of omega should be w/Hz or 1/time since Hz=(partial z)/(partial s) with units of length. I think this is consistent with the calculation of w in wvelocity.F

In the netCDF output file, omega is described as having units of meters/second,

Code: Select all

float omega(ocean_time, s_w, eta_rho, xi_rho) ;
	omega:long_name = "S-coordinate vertical momentum component" ;
	omega:units = "meter second-1" ;
	omega:time = "ocean_time" ;
	omega:grid = "grid" ;
	omega:location = "face" ;
	omega:coordinates = "x_rho y_rho s_w ocean_time" ;
	omega:field = "omega, scalar, series" ;

In varinfo.dat, omega is described as having units of m^3/s

Code: Select all

'omega' ! Output
'S-coordinate vertical momentum component'
'meter3 second-1' ! [m3/s]
'omega, scalar, series'
'ocean_time'
'idOvel'
'w3dvar'
1.0d0

My best guess, based on the comments and calculations in wvelocity.F is that the true units of omega should be 1/s, but I would like to confirm it has not been further modified (and that I am not being silly) somewhere between wvelocity.F and the output.

Oops, I just found this post: viewtopic.php?t=2139 . Krystin Thyng tracks down that omega is multiplied by Hz before being written out, so has units of m/s; can we change this in varinfo.dat?

Jamie

[jcwarner]

i think omega is written to the his file as m/s. here is what i see:

in omega.F, W (omega) is computed as
W(i,j,k)=W(i,j,k-1)- &
& (Huon(i+1,j,k)-Huon(i,j,k)+ &
& Hvom(i,j+1,k)-Hvom(i,j,k))
so it has units of m3/s

in wrt his, we have
!
! Write out S-coordinate omega vertical velocity (m/s).
!
CALL scale_omega (ng, tile, LBi, UBi, LBj, UBj, 0, N(ng), &
& GRID(ng) % pm, &
& GRID(ng) % pn, &
& OCEAN(ng) % W, &
& Wr3d)

scale omega is at the bottom of omega.F and does
Wscl(i,j,k)=W(i,j,k)*pm(i,j)*pn(i,j)
which makes Wscl = m3/s *1/m * 1/m = m/s
then Wr3d=Wscl

back to wrt his:
status=nf_fwrite3d(ng, iNLM, HIS(ng)%ncid, HIS(ng)%Vid(idOvel), &
& HIS(ng)%Rindex, gtype, &
& LBi, UBi, LBj, UBj, 0, N(ng), scale, &
& Wr3d)
and that is writing out the scaled Wscl as m/s.

Looking at Hernan's notes:
http://www.myroms.org/images/b/b22.jpg
http://www.myroms.org/images/b/b23.jpg

he derives the equation shown in wvelocity.F
! Hz * omega = w - d(z)/d(t) - div(z)
but the code writes out omega as m/s, not just /s.

-j

[jpringle]

Yes, I agree. All that is left is to change the text in varinfo.dat to reflect this.
Jamie

[arango]

Well, the omega vertical velocity in ROMS is special and its orthogonality changes at each grid point because it depends on the bathymetry. It does not have a physical interpretation since it is only related to ROMS descrete, terrain-following coordinates, 3D continuity equation. Internally in ROMS, it has units of transport (m³/s) because that's how it used in the flux-form, right-hand-side terms. We wanted to avoid repetitive multiplications. See ROMS governing equations. Also, there is the issue of Hz in the predictor/corrector time-stepping that depends on zeta. The value of Hz changes in the kernel since we do have time indices (nrhs, nstp, nnew) for it. So, we have specific reasons on how we time averaged omega in the averages output NetCDF file. On output, it is writen in m/s so we can compare with the true vertical velocity.

If you need true vertical velocity outside of ROMS, you need to activate the writing of such a variable into the output history NetCDF file, Hout(idWvel).

[jpringle]

Hernan --

I agree with all that. All I am saying is that varinfo.dat should be changed to be consistent with the actual units written out.

Jamie

[arango]

The varinfo.dat is an user-changeable file. In the beginning, we were thinking of writing omega in the restart NetCDF file as m³/s, and true vertical velocity in history files. But that's no longer the case. Nowadays, omega is recomputed internally. I rarely activate the switch for writing omega in my applications. I just do it when checking the wvelocity algorithm and comparing with true vertical velocity in flat bathymetry.

However, notice that in the output history NetCDF file omega have the correct units:

Code: Select all

        double w(ocean_time, s_w, eta_rho, xi_rho) ;
                w:long_name = "vertical momentum component" ;
                w:units = "meter second-1" ;
                w:time = "ocean_time" ;
                w:standard_name = "upward_sea_water_velocity" ;
                w:grid = "grid" ;
                w:location = "face" ;
                w:coordinates = "x_rho y_rho s_w ocean_time" ;
                w:field = "w-velocity, scalar, series" ;
        double omega(ocean_time, s_w, eta_rho, xi_rho) ;
                omega:long_name = "S-coordinate vertical momentum component" ;
                omega:units = "meter second-1" ;
                omega:time = "ocean_time" ;
                omega:grid = "grid" ;
                omega:location = "face" ;
                omega:coordinates = "x_rho y_rho s_w ocean_time" ;
                omega:field = "omega, scalar, series" ;

Anyway, your point is well taken. Thank you.

[jpringle]

Thanks