High resolution domain fail when loading VAR_SSO

[yairsuari]

Hello everyone and thanks in advance
I am trying to build one of my first domains. It has 4 nested domain (namelist.wps attached).
WRF initiation fails with the following error (quated from the rsl.0000):

Timing for processing wrfinput file (stream 0) for domain 4: 0.28118 elapsed seconds
-------------- FATAL CALLED ---------------
FATAL CALLED FROM FILE: <stdin> LINE: 268
topo_wind requires VAR_SSO data
-------------------------------------------

This happens only on domain 4 my grid spacing is (9,3,1,0.3)
I cant figure this out, as far as I can see, VARSSO dataset exists (found it in the data folder), so I was suspecting inconsistency in GEOGRID.TBL (also attached). I am using wrf using the uems interface so I am not certain this TBL file is the actual one I use and attach an additional TBL (GEOGRID.ARW.TBL).

can someone please help me crack this?
Thanks!

 

[kwerner]

Hi,
Can you take a look at your geo_em* files and see if you have VAR_SS0 data on all 4 domains? Thanks!

 

[yairsuari]

Thanks for helping.
I think its there (here is the ncdump output for the crashing domain):

ncdump -h geo_em.d04.nc | grep SSO
float VAR_SSO(Time, south_north, west_east) ;
VAR_SSO:FieldType = 104 ;
VAR_SSO:MemoryOrder = "XY " ;
VAR_SSO:units = "meters2 MSL" ;
VAR_SSO:description = "Variance of Subgrid Scale Orography" ;
VAR_SSO:stagger = "M" ;
VAR_SSO:sr_x = 1 ;
VAR_SSO:sr_y = 1 ;
:FLAG_VAR_SSO = 1 ;

And for another domain

ncdump -h geo_em.d03.nc | grep SSO
float VAR_SSO(Time, south_north, west_east) ;
VAR_SSO:FieldType = 104 ;
VAR_SSO:MemoryOrder = "XY " ;
VAR_SSO:units = "meters2 MSL" ;
VAR_SSO:description = "Variance of Subgrid Scale Orography" ;
VAR_SSO:stagger = "M" ;
VAR_SSO:sr_x = 1 ;
VAR_SSO:sr_y = 1 ;
:FLAG_VAR_SSO = 1 ;

Thanks again!

 

[kwerner]

Hi,
Can you send the following?
1) namelist.input
2) A met_em* file for each domain (these may be too large to attach. If so, see the home page of this forum for instructions of sending large files)
3) Please attach the full error log.
4) Let me know the version of WRF and WPS you're using
5) Let me know if you've modified the code in any way.

Thanks!

 

[yairsuari]

Thanks for helping!
I attach The requested files (I am using UEMS V 21.1 which works with WRF and WPS 4.3.3).
I am not there yet in terms of touching the code. However I installed WRF through UEMS (https://strc.comet.ucar.edu/software/uems/) and it is possible (but unlikely) that some manipulation of the WRF code was done.
My met_em files could not be attached so I uploaded them to Nextcloud but I was not able to find the link. (unless this is the link:https://nextcloud.mmm.ucar.edu/nextcloud/index.php/s/DkPamzUKYhQkWEI#)

Their names are:

met_em.d01.2022-03-09_00_00_00.nc
met_em.d02.2022-03-09_00_00_00.nc
met_em.d03.2022-03-09_00_00_00.nc
met_em.d04.2022-03-09_00_00_00.nc


Just as an additional inquiry, I tried to run this with a ratio of 1 between nesting 3 and 4 to see if this is really a resolution related problem and yet, domain 4 failed with the same error.
Thank you for your help
Yair

 

[yairsuari]

Maybe some additional data might help
I looked at the VAR_SSO on the failing domainmet_em file and found out that they are all zeros.
any guess as to what might be the reason?
a shift in the interpolation index file?
the resolution of the interpolated dataset is not high enough?
here is the line from namelist.wps

geog_data_res = 'srtm_3s+modis_30s+modis_lai+2m+default', 'srtm_3s+modis_30s+modis_lai+30s+default', 'srtm_3s+modis_30s+modis_lai+30s+default', 'srtm_3s+modis_15s+modis_lai+30s+default'

Thanks

 

[kwerner]

Thanks for sending those files. The reason you're not seeing VARSSO on d04 is because the values actually are all zero in that domain. If you take a look at VAR_SSO for domain 3, you see that there are only values >0 in the top left corner. Your domain 04 only overlaps d03 in areas where VAR_SSO is 0. You could turn off the topo_wind option for domain 4, and I believe it should work okay. Try that and let me know!

I also notice that in your namelist.wps you're using several different settings for the variable geog_data_res. I would recommend trying to simplify this a bit and perhaps if you're wanting to use your SRTM data, you can set each domain to
geog_data_res = 'srtm_3s+default'

and that would be sufficient. It may not be necessary, but if you're running into other problems, or if your results don't seem reasonable, I'd suggest trying that.

 

[yairsuari]

Thanks!
I somehow almost figured it out
Trying to rebuild the domain