problem of south china sea in coawst model

[limaolin]

I have run the south china sea in single roms successfully used the climate files.
Then I make the files for coawst model, when i run the south china sea in coawst model, there are some mistakes, thanks for help

Model Input Parameters: ROMS/TOMS version 3.7
Friday - March 25, 2016 - 7:07:46 PM
-----------------------------------------------------------------------------

Program received signal SIGSEGV: Segmentation fault - invalid memory reference.

Backtrace for this error:
#0 0x7F7B9410A4C7
#1 0x7F7B9410AB0E
#2 0x7F7B936119FF
#3 0x5645F4 in load_s2d_
#4 0x61C7BC in read_phypar_
#5 0x558DE8 in inp_par_
#6 0x40344D in __ocean_control_mod_MOD_roms_initialize
Segmentation fault (core dumped)

[kate]

#3 0x5645F4 in load_s2d_
#4 0x61C7BC in read_phypar_

This happens in the parsing of the ocean.in file. You can't use the same file for coawst as for the trunk ROMS - you have to use one with all the inputs that coawst is looking for. In particular, it needs:

Code: Select all

NCLMFILES == 1                         ! number of climate files
NBCFILES == 1                          ! number of boundary files

since it has the capability of reading more than one, just like the forcing files. Also, be sure that if you say you have 12 forcing files that you really have 12 forcing files.

[zhoulilong]

#3 0x5645F4 in load_s2d_
#4 0x61C7BC in read_phypar_

This happens in the parsing of the ocean.in file. You can't use the same file for coawst as for the trunk ROMS - you have to use one with all the inputs that coawst is looking for. In particular, it needs:

Code: Select all

NCLMFILES == 1                         ! number of climate files
NBCFILES == 1                          ! number of boundary files

since it has the capability of reading more than one, just like the forcing files. Also, be sure that if you say you have 12 forcing files that you really have 12 forcing files.

kate,I'd like to know how you confirm where the error happend from the code:
#3 0x5645F4 in load_s2d_
and
#4 0x61C7BC in read_phypar_
Is there a tool could translate "0x5645F4" code?

[kate]

How do I know this is what went wrong? You can confirm it by recompiling with USE_DEBUG=on, which should give you the line numbers in the .f90 files. However, I know that in the trunk code load_s2d gets called exactly once, on reading the forcing filenames. I know that in my branch (and coawst), it gets called three times. I know that I've spent time watching load_s2d in the debugger. After all, when I added the other calls to load_s2d, I didn't get it right the first time.

[limaolin]

Thanks all. I have solved this problem according to your advise.

Then I run it again, there is another mistake as follows:

CHECK_MULTIFILE - Error while processing Climatology multi-files:
data does not include initialization time = 56228 00:00:00

***** **:**:** ***** **:**:** Projects_My/scs_1/3D_BC/3_days/scs_1_clm.nc

But I check the time of the files:
the time of ini file is 56228
the time of bdy and clm files are 56228 56229
the time of frc file is 56228 to 56231

and Time-Stepping parameters as follows

NTIMES == 360 !1day
DT == 240.0d0 !4minutes
NDTFAST == 30 !2hour

how can I solved the problem, thanks!

[limaolin]

thanks all, I have solved the problems with your help.