[MITgcm-devel] EOS modification

[Jean-Michel Campin]

Hi Martin,

Concerning the EOS things, we were thinking of making a 
package with all the different Equation of states.

My view (I should say, "narrow" view) was to try to
make 1 subroutine for each EOS, that would return
either rho or alpha or beta (or 2 or the 3 at the same time).
Those EOS S/R can be called within find_rho & find_alpha/beta
and made relatively flexible with T & S as argument and 
also the size of the input/output arrays so that it can 
be called for 1 single grid point or 1 level of grid-points
(this would avoid the double version of each EOS).
I don't know precisely how it can be done, but my
impression was that it is feasible.
It goes in the same direction as the modification you suggest
(not calling find_rho in find_alpha/beta), and it would not be a 
problem for me that you make this modification.

But for now, it might be quicker just to fix this rhoConst problem
and not to change anything else so that Patrick does not have to check
again if TAF is happy or not with theses changes, since he will have 
do it anyway when the EOS package is finished.

Patrick, what do you think ?

See you,

Jean-Michel