[MITgcm-support] C grid and pressure

[Jean-Michel Campin]

Hi Nicolas,

On Fri, Jan 06, 2012 at 06:53:24PM -0500, Nicolas Grisouard wrote:
> Hi everyone,
> 
> Two simple questions, they should be easy but I'm having a hard time
> confirming my guesses.
> 
> 1/ In the pdf of the MITgcm manual, if you take a look at figure 2.9
> (the one showing the 3D staggering of velocity components), you see
> that w is evaluated on top and bottom of a tracer cell. So if you
> have Nr tracer cells, you should get Nr+1 vertical points for the W
> field, which is not the case. Does it mean that the bottom point is
> systematically ignored as is it zero? And hence that the
> z-coordinate for the W field is RF minus its bottom point?

You are right, the Nr+1 value is always identically zero and is not
stored in wVel array. And regarding the position, assuming the origin 
of the vertical axis (z=0) corresponds to the resting/flat (reference) 
sea-level, wVel(:,:,k) is just located at rF(k), =< 0. 

> 
> 2/ I am running oceanic, non-hydrostatic configurations with a
> linear free surface (nonlinFreeSurf = 0) and linear EOS. If I want
> to have the pressure at a given point i,j,k located in the center of
> a tracer cell, do I need to simply do the following:
> 
> P(i,j,k) = PH(i,j,k) + PNH(i,j,k) + Eta(i,j)*g*rho_surface(i,j) ?
> 
> PH, PNH and Eta are the result of reading the output files of the
> same names and rho_surface will be diagnosed from the temperature.
> What about PHL in the presence of topography? After searching
> through the documentation as well as the support list archive, I
> finally managed to get completely confused on that issue.