Ask Your Question
0

coupling WRF_CROCO

asked 2023-11-11 12:42:42 +0100

yxli gravatar image

updated 2023-11-23 03:02:46 +0100

Hello, I am having the following problem when running coupled WRF-CROCO. How can I solve it?

The error in running is in file 1.txt C:\fakepath\1.txt

I have attached some documents that may be required. C:\fakepath\mynamelist.txt C:\fakepath\submitjob.txt C:\fakepath\myjob.txt

I have read your code and steps carefully and I am not currently setting up CPLMASK nor running create_oasis_grids_for_wrf. I'm not sure if that's what caused the error. And the CROCO tutorials said "the CPLMASK creation may also be performed automatically in the coupling tools". I'm not sure exactly how to set it.

And The WRF I used was not supplied by croco.

==============================new============================================ I modified the netcdf path in jobcomp, now log.pile is as follows, does that mean there are no errors in this part? C:\fakepath\submitjob01.log And the following is the output log file for submitjob.sh. C:\fakepath\log.compil02.txt

C:\fakepath\out_run.txt

edit retag flag offensive close merge delete

3 answers

Sort by » oldest newest most voted
0

answered 2023-11-22 09:32:55 +0100

swen gravatar image

Coupling is very sensitive to netcdf libraries. I think that your nco is overwriting the paths to the original netcdf librarry you used to compile the models, and that's why you have an error. You need nco for preprocessing, but then you need to unload the nco module, and come back to the original netcdf library you used to compile when you want to run.

The toolbox is set up of that, but to use this functionality, you need to define your nco module in an environment variable named 'ncomod' in the myenv_mypath.sh file :

export ncomod='your_nco_module'

Then, the scripts are prepared to load this module when needed, and unload it after, to avoid any conflicts with the netcdf library that you need for running (same than that you compiled with).

So basically, you should have at the beginning of myenv_mypath.sh something that looks like this example (with your own modules):

!/bin/bash

#################################################################
################### FOR COMPILATIONS ############################
#################################################################

module purge
module load chpc/parallel_studio_xe/2020u1
module load chpc/earth/netcdf/4.7.4/intel2020u1

export PATH=/apps/chpc/earth/NetCDF-c461-f444-intel2020/bin:$PATH

export LD_LIBRARY_PATH=/apps/chpc/earth/NetCDF-c461-f444-intel2020/lib:$LD_LIBRARY_PATH
export LIBRARY_PATH=$LIBRARY_PATH:/apps/chpc/earth/NetCDF-c461-f444-intel2020/lib
export NETCDF=/apps/chpc/earth/NetCDF-c461-f444-intel2020
export NETCDF_CONFIG=${NETCDF}/bin/nf-config

export ncomod='chpc/earth/NCO/4.9.3'

# -- Option for job launching
export MPI_LAUNCH='mpirun'
export MPI_ext="-configfile"
edit flag offensive delete link more

Comments

Hello, I have also solved this problem by modifying jobcomp. I would like to know if there is any problem in the following steps. Since I cannot upload files, I have added them in the new reply below.

yxli gravatar imageyxli ( 2023-11-23 02:43:37 +0100 )edit
0

answered 2023-11-13 14:12:34 +0100

andres gravatar image

This kind of error messages

ncwa: command not found

indicate that you haven't installed the NCO package.

Please install it. In Ubuntu this is done with

sudo apt-get update sudo apt-get install nco

and try again.

edit flag offensive delete link more

Comments

Thank you for your advice. Since I was running the WRF-CROCO model in a cluster, I installed nco using conda. But after I load the nco environment, the new problem occurs.

yxli gravatar imageyxli ( 2023-11-21 08:39:05 +0100 )edit
0

answered 2023-11-21 06:35:44 +0100

yxli gravatar image

updated 2023-11-21 08:38:19 +0100

Since I was running the WRF-CROCO model in a cluster, I installed nco using conda. But after I load the nco environment, this problem occurs. I have attached all log.compil. Also, netcdf will load the nco dependent path '……/miniconda/env/nco/lib', and my modification of the netcdf path in myenv_mypath.sh does not work.

mod_oasis_coupler_mp_oasis_coupler_setup_': mod_oasis_coupler.F90:(.text+0xb2ed): undefined reference tonetcdf_mp_nf90_open_'

C:\fakepath\log.compile.txt

The following is the output log file for submitjob.sh. C:\fakepath\submitjob.log

=================================modified====================================== I modified the netcdf path in jobcomp, now log.pile is as follows, does that mean there are no errors in this part? C:\fakepath\log.compil02.txt And the following is the output log file for submitjob.sh. C:\fakepath\submitjob01.log

edit flag offensive delete link more

Comments

May I ask if mv a.out croco appears on/croco_in /log.compile does it mean there is no problem?

yxli gravatar imageyxli ( 2023-11-21 07:47:54 +0100 )edit
Login/Signup to Answer

Question Tools

1 follower

Stats

Asked: 2023-11-11 12:42:42 +0100

Seen: 27 times

Last updated: Nov 23