pyroms zslice vertical interpolation

[bilge.tutak]

Hi all,

In pyroms when you use quiver or zview in pyroms_toolbox, it ends up using pyroms.tool.zslice() function. And in the implementation of the zsclice it uses the FORTRAN implementation of zslice from iso.f in the "src" directory.

However, when I check the implementation of iso.zlice subroutine in FORTRAN it only seems be produce information (by interpolation) between bottom sigma layer depth and surface sigma layer depth. For example I can not use 0 (zero or close to zero depth) for surface layer

Code: Select all

            if ((zk(1).le.depths(j,i)).and.(depths(j,i).le.zk(N))) then
              call lintrp (N,zk,fk,1,depths(j,i),f2d(j,i))
            else
              f2d(j,i)=1.0d20
            endif

Is this really the case, or am I missing something?
Is there a way to get surface layer information?

Thanx

[kate]

In ROMS, tracers and other such values are valid between half a dz up from the bottom and half a dz down from the surface. That's what it's giving you. You could of course change the code so that the uppermost valid point is used all the way to the surface, but there's no new information there.

[bilge.tutak]

Thank you for the answer Kate;
The variables are defined at half dz from surface and bottom that is by the nature of the grid. However, what I meant was during visualization pyroms can only interpolate between top and bottom cells, which is logical. Alternatively, in John Wilkin's ROMS MATLAB tools, you can obtain results at points outside the bottom and surface cell depths. It is just a matter of interpolation/extrapolation. I just wanted to make sure about the situation in pyroms.

If you don't mind, in your applications, how do you obtain/plot surface values (temperature, velocity)?

Cheers,

[kate]

To get a quick look at the surface, I use ncview and look at the uppermost vertical level. Yes, it's half a dz down, but so what?

[bilge.tutak]

I agree, not much different information you can gain for half a dz up/down.

Thanks.