Custom Query (964 matches)

The code is okay, but I needed to make some modifications to the supplied TEST_HEAD input files to get bedload to be non-zero. These modifications are required because of the change in R224, where bedload was split from the other routines, and the looping was changed to DO ised=NCS+1,NST, yet the input files still used NCS=1, NNT=0 (NST=1), so the loop never gets executed.

I needed to modify the ocean_test_head.in file to set NCS=0, NNS=1, modify the sediment_test_head.in file to specify parameters for SAND instead of MUD, and also change the variable names from MUDFRAC_01 to SANDFRAC_01, MUD_01 to SAND_01, etc in the initialization NetCDF file test_head_ini.nc. Once I made these changes, the TEST_HEAD test case run with the R271 code (current trunk code) produced non-zero bedload values equivalent to versions prior to R224.

File ROMS/Utility/wrt_impulse.F is of the form

#include "cppdefs.h"
#if defined ADJOINT && defined IMPULSE
      SUBROUTINE wrt_impulse (ng, tile, model, INPncname)
...
      RETURN
      END SUBROUTINE wrt_impulse
#endif

Therefore it is empty if ADJOINT and IMPULSE are not both defined. This causes an error on some systems. One solution is to insert an empty subroutine in this case.

#include "cppdefs.h"
#if defined ADJOINT && defined IMPULSE
      SUBROUTINE wrt_impulse (ng, tile, model, INPncname)
...
      RETURN
      END SUBROUTINE wrt_impulse
#else
      SUBROUTINE wrt_impulse
      END SUBROUTINE wrt_impulse
#endif

In ROMS/Utility/get_bounds.F line 169

integer, intent(in)

Lfullgrid

should be

logical, intent(in)

Lfullgrid