READ_PHYPAR - Invalid dimension parameter

[manhnt181]

Hi everyone,
I got the error when I ran the COAWST model with the nested grids of ROMS and SWAN model. The detail of this issue is that:

Code: Select all

READ_PHYPAR - Invalid dimension parameter, NestLayers2
               must be equal to one in non-nesting applications.
 Found Error: 05   Line: 201      Source: ROMS/Utility/inp_par.F
 Found Error: 05   Line: 122      Source: ROMS/Drivers/nl_ocean.h, ROMS_initialize

I also attach all of the log files and the controlling scripts in the below.
Please help me to address this problem. I'm really appreciated that.
Thanks all,

[jcwarner]

this is for ROMS.
It looks like your ocean.in has

Ngrids = 2
NestLayers = 2

but it must have been compiled without nesting. the error is saying that you can only have NestLayers = 1 for non-nesting.
think you need to add
#define NESTING
to *.h file and recompile.
-j

[manhnt181]

Hi John,
Thank your for your reply.
You are right, I forgot using the #define Nesting option in the *.h project. According to your suggestion, I have used the #define Nesting option and recompiled model, this issue didn't appear again.
However, I got another error like that:

Code: Select all

forrtl: severe (174): SIGSEGV, segmentation fault occurred
Image              PC                Routine            Line        Source             
coawstM            0000000003C3E9CA  Unknown               Unknown  Unknown
libpthread-2.17.s  00002B5D9CF495D0  Unknown               Unknown  Unknown
coawstM            0000000000BF42A4  swboun_                  3546  swanpre2.f90
coawstM            0000000000AFE58E  swread_                  1851  swanpre1.f90
coawstM            0000000000A823FD  swan_initialize_          332  swanmain.f90
coawstM            0000000000A1EB75  waves_control_mod          94  waves_control.f90
coawstM            000000000041EFFD  MAIN__                    360  master.f90
coawstM            0000000000419122  Unknown               Unknown  Unknown
libc-2.17.so       00002B5D9D67E495  __libc_start_main     Unknown  Unknown
coawstM            0000000000419029  Unknown               Unknown  Unknown
Fatal error in MPI_Send: Other MPI error, error stack:
MPI_Send(173)..................: MPI_Send(buf=0x478ff10, count=4, MPI_INTEGER, dest=40, tag=13, comm=0x84000007) failed
MPIDI_EagerContigShortSend(262): failure occurred while attempting to send an eager message
MPIDI_CH3_iStartMsg(36)........: Communication error with rank 40
Fatal error in PMPI_Bcast: Other MPI error, error stack:
PMPI_Bcast(1478)......................: MPI_Bcast(buf=0x478f8c4, count=1, MPI_INTEGER, root=0, comm=0x84000004) failed
MPIR_Bcast_impl(1321).................: 
MPIR_Bcast_intra(1119)................: 
MPIR_Bcast_scatter_ring_allgather(962): 
MPIR_Bcast_binomial(213)..............: Failure during collective
MPIR_Bcast_scatter_ring_allgather(955): 
MPIR_Bcast_binomial(145)..............: 
dequeue_and_set_error(596)............: Communication error with rank 0

The line 3546 of the swanpre2.f90 program is that:

Code: Select all

IF (OPTG.NE.5) THEN                                         40.80
                 IY = CURR%JY                                             40.31
                 XC2 = XCGRID(IX,IY)
                 YC2 = YCGRID(IX,IY)

The line 1851 of the swanpre1.f90 program is:

Code: Select all

CALL SWBOUN ( XGRDGL, YGRDGL, KGRPGL, XYTST, KGRBGL, ng)

the line 332 of the swanmain.f90 program is:

Code: Select all

CALL SWAN_INITIALIZE (ng, ngc, ngp, NUM_SGRIDS, MyCOMM, Wname(ng))

The detail of the log files are attached in the below.
Please help me to solve this problem.
Thank you very much!

[jcwarner]

this is now a SWAN error.
I see in Errfile02-001 (this is for swan grid 2, tile partition 1)

Error : invalid boundary point
Error : ( -99 -99) is outside computational grid
Error : invalid boundary point
Error : ( -99 -99) is outside computational grid
Error : File is already opened: Projects/Tembin/TPAR1.txt
Severe error : unsupported boundary data file

so somehow the swan grid 2 has a boundary error.
For swan grid 2, you probably do not need any boundary data, as it should be getting all the data from the swan parent.

[manhnt181]

Dear Dr. John,
Thank you for your clear explanation.
Just like you said my error came from the the boundary of swan grid 2. After I clear out all of the boundary data from the 2nd grid of swan model, my model can run successfully.
Thank you very much!

[manhnt181]

Dear Dr. John,
I have another question related to use the create_roms_child_init and create_roms_child_clm programs. For example, when I need to create the ini, clm conditions of the 2nd grid, I use the ROMS grid and the ini, clm conditions from the parent grid. So when I run model with 3 nested domains, to generate the ini, clm conditions of the 3rd grid I have to use the 2nd ROMS grid and the ini,clm conditions, aren't I?
In the other hand, I need two contact files with the 2nd and the 3rd grid to run model, right? Similar with above question, the parent grid and the 2nd grid to create the contact file of the 2nd gridestion, I use the 2nd grid and the 3rd grid to create the contact file of the 3rd grid, right?
Can you kindly explain this issue for me?
Thank you very much!

[jcwarner]

half right, half not.
for ini and clm: you can use the parent ini and clm to make the first child ini and clm. Then use those first child ini and clm to make the files for the 3rd grid.
for the contact: roms only wants 1 contact file, so to create a contact file for 3 grids you need to make all 3 grids and then do something like:

Gnames={'roms_grid1.nc','roms_grid2.nc','roms_grid3.nc'}; % use the real grid names here
[S,G]=contact(Gnames,'roms_contact_for3grids.nc');

[manhnt181]

Dear Dr. John,
Thank you for your explanation. Based on your advices, I can generate the ini, clm conditions and the contact file as well.
However, during testing process, I got the problem in the COAWST.out file is that:

Code: Select all

 TIME-STEP YYYY-MM-DD hh:mm:ss.ss  KINETIC_ENRG   POTEN_ENRG    TOTAL_ENRG    NET_VOLUME  Grid
                     C => (i,j,k)       Cu            Cv            Cw         Max Speed

         0 2021-12-21 00:00:00.00           NaN           NaN           NaN           NaN  01
                     (000,000,00)  0.000000E+00  0.000000E+00  0.000000E+00  0.000000E+00
 Found Error: 01   Line: 341      Source: ROMS/Nonlinear/main3d.F
 Found Error: 01   Line: 331      Source: ROMS/Drivers/nl_ocean.h, ROMS_run

And the report of the Errfile is:

Code: Select all

Terminating error: whitecapping is already set
Terminating error: whitecapping is already set

All of the log files are attached in the below. I'm so sorry because this error is not related to above problem.
Please help me to address it.
Thank you very much!

[jcwarner]

there are probably a few things not correct here, you will need to try and figure them out.

for swan Err file01 -001 : this is for grid 1, tile 1 (so probably left side)
it has Error : invalid boundary point
Error : ( -99 -99) is outside computational grid
so most likely one of the Boundspec locations is outside the grid. you may need to comment them out one by one to find out which one is outside the grid.

for roms: the NAN
could be from many things. look at the entire std out file, all of the lines above what you show. And see if there are any issues above (in the grid, in the forcings, etc, see if there are any nans in the forcings).

[manhnt181]

Dear Dr. John,
Thank you for your help.
After I checked my ini condition, I found out that the values of u, v and zeta variables = 0. Indeed the problem came from my ini file.
Thank you very much!

[manhnt181]

Dear Dr. John,
My problem came from the clm, ini, bdy conditions. All of the values of u,v, temp, zeta variables are zero. The data that I used for this case was in the GLBy0.08/latest dataset URL. I also tried using the datasets from GLBy0.08_930_FMRC URL with the 00Z and 12Z 26Dec2021, however, the issue still occurred.
Can you kindly help me to resolve this problem?
Thank you very much!

[zhaolv]

How did you resolve the clm to 0? The data I extracted from GLBa0.08 is also 0