[MITgcm-support] Potential density - ignore previous message

[Sergio Jaramillo]

Martin ,

Sorry about my last message, I was still too sleepy, the only thing that 
is true from it is the values of rho, rhonil, and tAlpha, that tRef is 
not what I use!. Now, what you say is:

rho_reference= rhoNil*(1+tAlpha*tRef(z)),


then the refernce density is proportional to the temperature?

what I was doing actually was using the linear equation of state 
(Turner,1973):

rho = rhoNil*(1+sBeta*S - tAlpha*T)

then, T= (rhoNil - rho)/(tAlpha*rhoNil);

and using this for the input tRef. Here density~-T, which was more 
intuitive for me. But I have that doubt about the tAlpha increased one 
order of magnitude just to have the model running.

Again sorry for the big mess.

Sergio