bottom vertical velocity

[simion1232006]

Hi everyone:

Is there anyone notice in ROMS output the vertical velocity at the bottom is not zero.
In my study, I tried to look at the tracer transports. I integrated the vertical heat flux through the depth and found it is not zero, so
I checked the vertical velocity from the history file, and found it is not zero along the water column. How could that happen? I mean the vertical velocity(w) both at surface and bottom should be zero ,otherwise it will conflict the Physical nature. The vertical velocity(omega) in sigma coordinate is correct though.

I need to point out that my algorithm to compute heat flux should be correct. if i manually changed vertical velocity to zeros on surface and bottom , the vertical integrated heat flux is always zero which is normal.

I really appreciate if someone can explain it to me.

[jcwarner]

omega is the flux thru the grid cell faces, and is perpendicular to the cell face. The w velocity is aligned with gravity, and thus may be rotated in reference to the cell face. IF your test case has a flat bottom and a flat water surface, then omega and w should be the same (there is also a scale factor of pm*pn here).
see ROMS/Nonlinear/wvelocity.F:
In ROMS, the terrain-following vertical velocity, omega, is given by:
!
! Hz * omega = w - d(z)/d(t) - div(z)
!
! where w is the "true" vertical velocity and
!
! div(z) = pm * u * d(z)/d(xi) + pn * v * d(z)/d(eta)

[simion1232006]

Thank you.Dr Warner.
My bathymetry is not flat.That should be the reason.

Can i ask another question:
If I calculate the heaflux from ROMS history file, how could I make sure the hearflux is conservative.
For each grid, i can compute the hearflux to the x-direvtion, y-direction,but for vertical direction, should i use oemga velocity or w?
As I found part of the vertical flux have already been computed in x-direction or y-direction(the grid is not flat) due to sigma coordinates.