Custom Query (986 matches)

It has been decided to restructure the routine sediment.F. That routine was becoming very lengthy and difficult to manage. The restructuring consists of distributing the components of sediment.F into several new routines and adding new routines. This will allow the different components to be more easily updated and to more easily allow future contributions.

Routines that were modified/created:

M      ROMS/External/ocean_shoreface.in
M      ROMS/External/coupling_inlet_test.in
M      ROMS/External/varinfo.dat
M      ROMS/Nonlinear/sediment.F
M      ROMS/Nonlinear/bedbiodiff.F
A      ROMS/Nonlinear/bed_cohesive.F
A  +   ROMS/Nonlinear/bottom.F
M      ROMS/Nonlinear/bedload.F
M      ROMS/Nonlinear/bed.F
A  +   ROMS/Nonlinear/dep_ero.F
M      ROMS/Nonlinear/settling.F
M      ROMS/Utility/inp_par.F
M      ROMS/Modules/mod_sediment.F
M      ROMS/Modules/mod_ncparam.F
M      ROMS/Modules/mod_ocean.F

These files are in my branch. Please review for inclusion to distributed code.

for the SWAN coupling, the call mpi_finalize is now located at the bottom of Master/mct_coupler.h

Here is the problem:

It is not clear which model will get there first. Typically this will be ROMS. However, SWAN needs time to collect the contents of the individual output files and conglomerate them into single output files (don't ask me why SWAN does this, it is just the way it is). But if ROMS goes thru it's finalize step, then calls mpi_finalize, well that is bad news for the SWAN because it may not have finished pulling together the output.

Suggested Solution:

Place an mpi_barrier in mct_coupler.h before the call to mpi_finalize so that all processors from all models need to check in before finalize occurs.

In ROMS/Utility/def_floats.F, line 334

              WRITE (Vinfo(19),20) 1000.0_r8*Sd50(i,ng)

should be

              WRITE (Vinfo(19),40) 1000.0_r8*Sd50(i,ng)

Corrected file attached.