[Mitgcm-support] towards chp37

[mitgcm-support at dev.mitgcm.org]

Hello,

I have check-in few changes related to the units of the
pressure/potential field. I run the test-experiments
and looks OK. But because of change in Matrix of implicit
solver, no "passed at highest accuracy" is expected.

I think its time to make a checkpoint.
Here I summarize the latest changes:
----

The objective was to allow a more precise surface-geopotential 
gradient for the atmosphere, especially with topography, taking 
into account the change in space of Alpha (=1/rho) near the surface.

Change in units: 
Phi_Hydrostatic (=phiHyd), Phi_Surface(=cg2d_x) and 
Phi_NonHydrostatic (=cg3d_x), all have now the dimension of a potential 
[m2/s2] (pressure/rhoConst in the ocean ; Geo-Potential in the atmosphere).

The global variable "eta" is the surface R-anomaly [r unit]
(ocean: sea surface height [m] / atmos: surface pressure anomaly [Pa])
and has been added (since checkpoint36).

Just to remember, in checkpoint36 and before:
Phi_Surf : cg2d_x in meter (ocean & atmos)
Phi_NH   : cg3d_x in meter (ocean)
Phi_Hyd  : phiHyd   ocean: Pressure ; atmos: GeoPot, m2/s2

Units are also consistent inside the 3D solver if P-coordinate is used.
For the 2 Implicit solvers (CG2D and CG3D), the changes affect the matrix
(ocean: previous/g) and consequently the normalization factor.
( Question: Do we needs to update "results/output.text" 
  of some (or all) test-experiments since cg2d_nom has changes ?)

Variable "gBaro" is no longer used.

See you,

Jean-Michel