In situ or potential density anomaly?

[agomer]

Hi everyone!

I am having a problem with the initial density profile using Roms Agrif.
My simulations are very idealized, so I define analytically the initial conditions. In particular, I define an exponential decaying temperature profile and a constant salinity profile (see plots attached), with the flags NONLIN_EOS and SPLIT_EOS activated. Then, if I output the density anomaly (rho1=rho-1025), it seems to be the potential density anomaly, rather then the expected in situ density anomaly!

The third figure shows some vertical density anomaly profiles that I have calculated with a couple of different methods:
* 'density' is the output from my simulation
* 'potential density (sw)' is the potential density anomaly calculated with the Seawater library for Matlab Version 3.2
* 'density (sw)' is the in-situ density anomaly calculated with the same library as above
* 'density (dcherian)' is the in-situ density anomaly camputed with this script: https://github.com/dcherian/tools/blob/ ... roms_eos.m

Has enyone ever had anything similar? Any help is very appreciated!
Agostino