No IMPLICIT type set for set_zeta_timeavg

[jake11230]

Hello I am needing some help getting ROMS to compile correctly. I was able to get the UPWELLING test case to compile correctly but now I am trying to use a header (.h file) that a former colleague had given to me for running ROMS to compute storm surge. When using this header file, the compiling fails with: Error: Symbol 'set_zeta_timeavg' has no IMPLICIT type. It is in the ini_fields.f90 file, but I'm not as fluent with fortran so not really sure what I need to do to correct this issue, so I was hoping to get some assistance if possible. I have attached the header file for reference, if anything else is needed let me know.

[wilkin]

To debug this I recommend you keep your working test case UPWELLING more or less as is, and then successively add features (#define ...) from stormsurge.h to work toward the dynamics configuration you seek. Then you can zero in on the specific feature that is triggering the problem.

[jcwarner]

can you type
grep -nir set_zeta_timeavg ROMS/*
and find out what folder that var is in?

[jake11230]

I'm working on adding in the #define features to the UPWELLING case and haven't gotten the same set_zeta_timeavg error yet. I imagine this is because the default UPWELLING case didn't trigger it, so there must be one of the features in there that are defining it already. I'm still working through some more of them though so I might be able to find the one that is causing the issue.

As for the grep suggestion, the output from that command is attached. The ROMS/Nonlinear/ini_fields.F is the file that is triggering the error but I'm not sure how to find which header feature is triggering the error

[jcwarner]

as i look thru the ini_fields, i see set_zeta_timeavg, but it is inside a solve3d loop, for example:

925 # ifdef SOLVE3D
926 !
927 !-----------------------------------------------------------------------
928 ! Initialize fast-time averaged free-surface (Zt_avg1) with the inital
929 ! free-surface
930 !-----------------------------------------------------------------------
931 !
932 CALL set_zeta_timeavg (ng, tile, model)
933

the only part that is not, is near the top.
PUBLIC :: set_zeta_timeavg (line 58)

maybe try
# ifdef SOLVE3D
PUBLIC :: set_zeta_timeavg
# endif

[jake11230]

Oh thanks for pointing that out. I added your suggestion to the code and that allowed it to compile without any errors. Thanks for your help getting this resolved.