Opened 17 years ago
Closed 17 years ago
#14 closed request (Invalid)
SWAN / ROMS coupling problem
| Reported by: | mathieu | Owned by: | arango |
|---|---|---|---|
| Priority: | major | Milestone: | Release ROMS/TOMS 3.1 |
| Component: | Coupling | Version: | 3.1 |
| Keywords: | Cc: |
Description
The output abilities of ROMS are superior to SWAN so naturally, one wants to output with ROMS in netcdf instead of SWAN in ascii file. So, one set up no output between the two, i.e. NREOQ=0
But SWOUTP the subroutine for output also determine the value of "cplout" for coupling with ROMS and it also realize the interfacing. So with NREOQ=0 no coupling.
On the other hand if NREOQ=4 (4 field outputed), ocean_coupling is called 4 times, which is not correct.
Solution is to move the coupling outside of the 70 loop of SWOUTP.
Change History (1)
comment:1 by , 17 years ago
| Component: | Nonlinear → Coupling |
|---|---|
| Resolution: | → invalid |
| Status: | new → closed |
| Type: | bug → request |
Note:
See TracTickets
for help on using tickets.
To run the coupled system, please use the INPUT_inlet_test or other INPUT file distributed with ROMS as an example. The user needs to have the output requests listed in those INPUT files. These output requests are flags to SWAN so that the necessary fields are computed. For example, wave height is not a prognostic variable in SWAN. SWAN needs to compute Hsig for output purposes and for the coupling. We tell SWAN to compute that variable by using the ouput request as listed in the INPUT file. Users can set that output to be very infrequent if they do not want it, but the command must be there so that SWAN computes it for the coupling.
In the future we may add the capabilites to have other variables from SWAN written thru ROMS into the NetCDF files. We are formulating a master coupling plan to deal with multiple models and multpile needs for similar data.