About the psi-variable

Discussion of how to use ROMS on different regional and basin scale applications.

Moderators: arango, robertson

Post Reply
Posts: 8
Joined: Mon Sep 28, 2009 6:08 pm
Location: Geosciences, Univ. of Oslo

About the psi-variable

#1 Post by eivinds » Wed Oct 28, 2009 12:59 pm

I am setting up the model with grid from file. My collegue assisted me in this, and he said that I could simply delete all variables related to psi in the input netcdf files, i.e x_psi, mask_psi etc. This way the model would not compute the psi values, resulting in faster run time and less model output. In the forum topic "Sediment for beginner))" Kate says that psi is a variable "..from a previous generation of the model." which support the view that psi is not a variable I should need to take into consideration.

When I try to run the model, I get the error:
"GET_GRID - unable to find grid variable: mask_psi in grid NetCDF file: vanm_grid.nc"

When looking in the file get_grid.F I see that the routine indeed looks for all the psi-related variables that I thought were unimportant. I suppose what I wonder is wether there is a way to avoid performing the psi-computations, and if so, how? If not its not a problem for me to simply put the psi-variables back in.


User avatar
Posts: 3682
Joined: Wed Jul 02, 2003 5:29 pm
Location: IMS/UAF, USA

Re: About the psi-variable

#2 Post by kate » Wed Oct 28, 2009 5:00 pm

For the most part, ROMS reads in what it needs and ignores the rest. It does use the mask at psi points, but insists on recomputing it from the rho mask, so it probably doesn't have to read it from the grid file first.

Psi is indeed not a ROMS variable, but there are a few computations that take place on what we call the "psi points".

Post Reply