Using LCZ map via w2w caused wrf.exe to terminate

[hwangjiacheng]

Describe your issue

Hi, when I was using the w2w tool, the pre-processing should have been done correctly. But the wrf.exe program broke at the beginning (after generating wrfout_d01, wrfout_d02). The rsl.error file has no obvious error message, only the following message is displayed on the operation interface. The version of wrf is 4.3.3, these are parameter files
namelist.input.txt
namelist.wps.txt
. The lcz map i use is from https://zenodo.org/record/7324909. How can we solve this problem?

w2w --version

0.5.0

nc-config --all

netcdf-c-4.7.2
netcdf-cxx4-4.3.0
netcdf-fortran-4.5.2

Installed Packages

Package         Version
--------------- ------------
affine          2.4.0
attrs           23.1.0
certifi         2023.7.22
cftime          1.6.2
click           8.1.7
click-plugins   1.1.1
cligj           0.7.2
ecmwflibs       0.5.3
findlibs        0.0.5
h5netcdf        1.2.0
h5py            3.9.0
netCDF4         1.6.4
numpy           1.25.2
packaging       23.1
pandas          2.1.0
pip             23.2.1
pyparsing       3.1.1
pyproj          3.6.0
python-dateutil 2.8.2
pytz            2023.3.post1
rasterio        1.3.8
rioxarray       0.15.0
scipy           1.11.2
setuptools      68.0.0
six             1.16.0
snuggs          1.4.7
tqdm            4.66.1
tzdata          2023.3
w2w             0.5.0
wheel           0.38.4
xarray          2023.8.0

Traceback

No response

[matthiasdemuzere]

Hi @hwangjiacheng,

• did you check the geo_em files provided by W2W? Do they look as expected?
• Which of the files do you then use as input for the model? The geo_em.dXX_LCZ_params.nc one?
• As far as I can see, your namelist.input.txt file looks ok?

If all of the above are ok, then it seems to be an issue on the wrf-side? @andreazonato, have you seen this before?

[andreazonato]

Hi @hwangjiacheng

Actually you gave me too few information to debug.

Can you please increase the debug_level in WRF?

And...did you check if real provides correct wrfinput?
And...you should turn on sf_urban_physics=3 for all the domains

[hwangjiacheng]

Hi @hwangjiacheng,

• did you check the geo_em files provided by W2W? Do they look as expected?
• Which of the files do you then use as input for the model? The geo_em.dXX_LCZ_params.nc one?
• As far as I can see, your namelist.input.txt file looks ok?

If all of the above are ok, then it seems to be an issue on the wrf-side? @andreazonato, have you seen this before?

Thank you for your reply.

• Yes, I have checked these files and they seem to be right? geo_em files.zip
  (below is the geo_em.d03 file)
  

• Yes, I used the geo_em.d03_LCZ_params.nc file.

• Here are some error logs that may provide more information.
  rsl.error.0000 (real.exe).txt
  rsl.error.0000 (wrf.exe).txt
  

[hwangjiacheng]

Hi @hwangjiacheng

Actually you gave me too few information to debug.

Can you please increase the debug_level in WRF?

And...did you check if real provides correct wrfinput? And...you should turn on sf_urban_physics=3 for all the domains

Thank you for your reply.

• Here are some files produced by the wrf that may provide more information. (wrfinput_d01 and wrfinput_d02 are too large to upload)
  geo_em files.zip
  rsl.error.0000 (real.exe).txt
  rsl.error.0000 (wrf.exe).txt
  wrfinput_d03.txt

• Thanks for your suggestion, I will try to raise debug_level. But I don't know about this parameter, what does it control?

• The error log of the real.exe shows that it is running correctly. But I'm just starting to learn wrf and don't know how to check wrfinput files.

• Thanks for your suggestion, i tried to turn on sf_urban_physics=3 for all the domains, but wrf.exe still couldn't run.

[andreazonato]

Quick question: is it running without LCZS?

[hwangjiacheng]

Quick question: is it running without LCZS?

Yes, it works. If using lczs, wrf.exe will stop after generating wrfout_d01 and wrfout_d02 for the first hour (wrfout_d03 cannot be generated).

[matthiasdemuzere]

@hwangjiacheng just checking here. Did you manage to resolve this issue?

[hihihiccc]

@hwangjiacheng have you managed to solve this problem yet? I've encountered a similar issue and haven't been able to resolve it so far.

[LJroy1998]

I am also facing the same issue. If you have already solve the issue @hwangjiacheng please let me know.

[andreazonato]

Which WRF versions are you running guys? Is it requiring LCZ to be 30-40 or 50-60?

[hwangjiacheng]

@LJroy1998 I changed to a better computer and the problem was solved. Could it be related to memory performance?

[hwangjiacheng]

@hihihiccc I changed to a better computer and the problem was solved. Could it be related to memory performance?

[LJroy1998]

@andreazonato i am using WRF version 4.3.

[LJroy1998]

@hwangjiacheng That could be possible i have to check

[andreazonato]

Maybe you should simply add processors to your computation. If you are using WRF version 4.3, then LCZ should go from 30 to 40

[LJroy1998]

I will look into that solution @andreazonato and let you know if it works. Thank you for the help

[Lagnajee]

@hwangjiacheng what is the configuration of your upgraded system?

[LJroy1998]

@hwangjiacheng can you please tell me what is the configuration of your upgraded computer?