t3dbc_im.f90 error

[nma]

Dear friends,

Recently I could see some errors in t3dbc_im.f90, which I could solve following a suggestion by Kate viewtopic.php?f=14&t=1229 by commenting out USE mod_setting. Further there is another error in the same routine given as follows:

cd /m/raid2/nma/Original_coawst/COAWST/Build; /m/raid2/nma/libs/test/mpich2-1.4/MPICH2/bin/mpif90 -c -frepack-arrays -O3 -ffast-math -I/m/raid2/nma/libs/test/MCT/buildMCT/include t3dbc_im.f90
t3dbc_im.f90:244.13:

t2=nnew(ngp)
1
Error: Function 'nnew' at (1) has no IMPLICIT type
t3dbc_im.f90:60.23:

& nstp(ng), nout, &
1
Error: Function 'nstp' at (1) has no IMPLICIT type
make: *** [/m/raid2/nma/Original_coawst/COAWST/Build/t3dbc_im.o] Error 1


Has anybody seen this before....
thanks and regards, nilima

[jcwarner]

this is my modeling system with the grid refinement.
The ngp (parent grid) time levels are used to determine the placement of data for the interpolation. I use nnew and nstp to set the pointer for the data. These flip back-forth: 2-1, 1-2, 2-1 etc. I had recently made some changes to this. Are you running a test case with the refinement? For now on, it may be best not to post these issues on the Rutgers forum so that users dont get confused. If you are running a refined grid test case, please do an update and send me an email if you are having troubles.
thanks,
john

[nma]

Yes John, I was doing a refined grid test case, where I got this error. Since the previous errors related to the same routine were posted here, so I thought it could be related to generic ROMS model error!
I shall update you on refined grid applications...
thanks and regards, nilima