initial and boundary conditions

[asujrpv]

Hello everyone,

I apologize if my questions are too basic, I am getting acquainted with ROMS and have little experience with it. I do appreciate your help in clarifying the following:


I have created the grid and bathymetry using seagrid. I now need to create the initial and boundary conditions for my specific problem.

I have downloaded the matlab and fortran tools developed by Hernan Arango in the initial.tar file.

The temperature and salinity profiles that I have for the region were taken on Sep 3, 2006 at a depth ~ 40 m, about 15 km to the west of a prominent bathymetric rise (the summit is ~ 25 m) seen in the middle of the domain.

I was thinking to take this T(z) and S(z) and use it as initial conditions at every point of the grid trimming them where the depths are shallower and linearly extending down to the bottom when they are deeper.

Is this possible to do using the fortran tools? I would need to put the temperature and salinity profiles in a netcdf format to be read by the program, is this correct?


If ROMS provides a database for the region of interest and I can find there T and S profiles similar to what we measured, maybe I can use those profiles to initialized the model.

The boundary conditions at the solid boundaries (mainly at the seafloor) should be specified in the code. What are the and I think that ROMS provides several choices.

Where or how do I select free-shear on top and non-reflective (open) boundary conditions?

I would be most grateful for your help.

Thanks!
Rafael

[kate]

I have downloaded the matlab and fortran tools developed by Hernan Arango in the initial.tar file.

The temperature and salinity profiles that I have for the region were taken on Sep 3, 2006 at a depth ~ 40 m, about 15 km to the west of a prominent bathymetric rise (the summit is ~ 25 m) seen in the middle of the domain.

I was thinking to take this T(z) and S(z) and use it as initial conditions at every point of the grid trimming them where the depths are shallower and linearly extending down to the bottom when they are deeper.

Is this possible to do using the fortran tools? I would need to put the temperature and salinity profiles in a netcdf format to be read by the program, is this correct?

Yes, you need to make netcdf files for initial conditions. You can start with the matlab tools and modify them to be using your T(z) and S(z). It is assumed that you have some proficiency with matlab or the similar tool of your choice for doing this kind of operation. I'm working towards that with python... You need to know the mapping between z and the ROMS s coordinate in order to do these things - these functions exist for matlab and python.

If ROMS provides a database for the region of interest and I can find there T and S profiles similar to what we measured, maybe I can use those profiles to initialized the model.

ROMS doesn't provide these, but we do have tools for extracting initial conditions from SODA or Levitus. Perhaps you know of something in your region.

The boundary conditions at the solid boundaries (mainly at the seafloor) should be specified in the code. What are the and I think that ROMS provides several choices.

Where or how do I select free-shear on top and non-reflective (open) boundary conditions?

So you want solid walls on the sides? What are you doing on the sides?

For the top and bottom, try these:

Code: Select all

** ANA_SMFLUX          use if analytical surface momentum stress
** ANA_SSFLUX          use if analytical surface salinity flux
** ANA_STFLUX          use if analytical surface net heat flux
** ANA_BSFLUX          use if analytical bottom salinity flux
** ANA_BTFLUX          use if analytical bottom temperature flux

and one of the bottom drag options.

[asujrpv]

I have downloaded the matlab and fortran tools developed by Hernan Arango in the initial.tar file.

The temperature and salinity profiles that I have for the region were taken on Sep 3, 2006 at a depth ~ 40 m, about 15 km to the west of a prominent bathymetric rise (the summit is ~ 25 m) seen in the middle of the domain.

I was thinking to take this T(z) and S(z) and use it as initial conditions at every point of the grid trimming them where the depths are shallower and linearly extending down to the bottom when they are deeper.

Is this possible to do using the fortran tools? I would need to put the temperature and salinity profiles in a netcdf format to be read by the program, is this correct?

Yes, you need to make netcdf files for initial conditions. You can start with the matlab tools and modify them to be using your T(z) and S(z). It is assumed that you have some proficiency with matlab or the similar tool of your choice for doing this kind of operation. I'm working towards that with python... You need to know the mapping between z and the ROMS s coordinate in order to do these things - these functions exist for matlab and python.

If ROMS provides a database for the region of interest and I can find there T and S profiles similar to what we measured, maybe I can use those profiles to initialized the model.

ROMS doesn't provide these, but we do have tools for extracting initial conditions from SODA or Levitus. Perhaps you know of something in your region.

The boundary conditions at the solid boundaries (mainly at the seafloor) should be specified in the code. What are the and I think that ROMS provides several choices.

Where or how do I select free-shear on top and non-reflective (open) boundary conditions?

So you want solid walls on the sides? What are you doing on the sides?

For the top and bottom, try these:

Code: Select all

** ANA_SMFLUX          use if analytical surface momentum stress
** ANA_SSFLUX          use if analytical surface salinity flux
** ANA_STFLUX          use if analytical surface net heat flux
** ANA_BSFLUX          use if analytical bottom salinity flux
** ANA_BTFLUX          use if analytical bottom temperature flux

and one of the bottom drag options.

Many thanks for the reply. What I want are open boundary conditions the bottom but on the domain boundary (non-reflection waves) - shear free on the top, non-penetration at the bottom and open on the sides.

Regarding input of T and S from ascii file to a netcdf file, maybe I can take the netcdf files in the examples to see how the variable is written?

I very much appreciate your patience and guidance.
Rafael