search

DOI-USGS / COAWST

COAWST modeling system git repository
Other
103 stars 48 forks source link

A problem about running coawst (qsub run_coawst) #55

Open LeoSoudelor opened 3 years ago

LeoSoudelor commented 3 years ago

Hi John,

I have finished setting the case I want to simulate following the user manual (only WRF-ROMS model, no SWAN). When I ran run_coawst, two problems occurred and showed on cwst.out. The first problem is about "NAT". I tried to find "NAT" error in two files in the below figure1, but I had no idea to deal with it. The second problem which is about MPI is show in below figure 2. Please help me deal with these problems, thanks! Figure1:

cwst out_NAT

Figure2:

cwst out_MPI

Files that may be required: coupling_HIGOS.txt cwst_out.txt namelist.txt ocean_HIGOS.txt run_coawst.txt

jcwarner-usgs commented 3 years ago 
     NAT =  30             ! Number of active tracers (usually, 2)

set NAT =1 if just temperature (temp is always on), and you should set NAT=2 if you have #define SALINITY.

for the mpi, not sure. that is just saying mpi stopped. i thik roms did not run because of invalid nat.

LeoSoudelor commented 3 years ago

Thanks for replying me so quickly! After modification, the output of cwst_out looks more reasonable. But I don't know why it stop at "DYNAMICS OPTION: Eulerian Mass Coordinate alloc_space_field: domain 1......"(as the follow cwst_out.txt) without showing 'error' ? cwst_out.txt

jcwarner-usgs commented 3 years ago

ok. are there any other files created? like a swan Err* file?

LeoSoudelor commented 3 years ago

Oh I find that except cwst.out , HIGOS_o75427 is also be created. But there are many Unknown in 'HIGOS_o75427'...... HIGOS_o75427.txt

jcwarner-usgs commented 3 years ago

couple of things, looking at your ocean_* .in file -need to add the tracer advection option, near the top, look in the sandy_ocean.in as an xample.

-change to this (remove the !, sometimes it just likes to find the information) : NUDNAME == ocean_nud.nc SSFNAME == ocean_rivers.nc

-change to NFFILES == 1 ! number of unique forcing files FRCNAME == Projects/Florence/Florence_roms_forcing.nc

if that still not work, then compile with DEBUG=on and run coawstG

LeoSoudelor commented 3 years ago

I have changed "USE_DEBUG=on" in "coawst.bash", and ./coawst.bash. But it arise errors like following figure. (This error didn't appear in my previous compile process) compile_error

atm2ocn_flux.f90(58): error #6631: A non-optional actual argument must be present when invoking a procedure with an explicit interface. [LINE] CALL wclock_on (ng, iNLM, 17) -----------^ atm2ocn_flux.f90(58): error #6631: A non-optional actual argument must be present when invoking a procedure with an explicit interface. [ROUTINE] CALL wclock_on (ng, iNLM, 17) -----------^ atm2ocn_flux.f90(78): error #6631: A non-optional actual argument must be present when invoking a procedure with an explicit interface. [LINE] CALL wclock_off (ng, iNLM, 17) -----------^ atm2ocn_flux.f90(78): error #6631: A non-optional actual argument must be present when invoking a procedure with an explicit interface. [ROUTINE] CALL wclock_off (ng, iNLM, 17) -----------^ compilation aborted for atm2ocn_flux.f90 (code 1) make: [Build/atm2ocn_flux.o] Error 1 make: Waiting for unfinished jobs....

jcwarner-usgs commented 3 years ago

i just pushed a new version of atm2ocn_flux.F. can you get that (it is in ROMS/Nonlinear)

LeoSoudelor commented 3 years ago

I have renew the Nonlinear file. And I ./coawst.bash again, but another two errors arised.(like the following figure)

螢幕擷取畫面 2021-08-06 020824

atm2ocn_flux.f90(72): error #6460: This is not a field name that is defined in the encompassing structure. [STFLUX] & FORCES(ng) % stflux, & -------------------------------------------^ atm2ocn_flux.f90(72): error #6634: The shape matching rules of actual arguments and dummy arguments have been violated. [STFLUX] & FORCES(ng) % stflux, & -------------------------------------------^ compilation aborted for atm2ocn_flux.f90 (code 1) make: *** [Build/atm2ocn_flux.o] Error 1

jcwarner-usgs commented 3 years ago

can you post your .h file

LeoSoudelor commented 3 years ago

These are .h files in ROMS/Nonlinear, thanks. h_files.zip

jcwarner-usgs commented 3 years ago

sorry. i meant your project.h file that lists your model settings.

LeoSoudelor commented 3 years ago

OK, sorry. This is my higos.h in /Project/HIGOS/. And because I run WRF+ROMS only (no SWAN), some setting is different from the user manual. higos_h.txt

LeoSoudelor commented 3 years ago

OK, sorry. This is my higos.h in /Project/HIGOS/. And because I run WRF+ROMS only (no SWAN), some setting is different from the user manual. higos_h.txt

Is there any problem of higos_h.txt ?