observation as ./ob.ascii is not reading in WRFDA. However, obs_gts.3dvar was successful.

[priya_nit]

./da_wrfvar.exe error:
module_io_quilt_old.F 2931 T

Namelist logging not found in namelist.input. Using registry defaults for variables in logging.

*** VARIATIONAL ANALYSIS ***
WRFDA V4.5.2

Ntasks in X 10, ntasks in Y 10
*************************************
Parent domain
ids,ide,jds,jde 1 304 1 294
ims,ime,jms,jme -4 38 -4 37
ips,ipe,jps,jpe 1 31 1 30
*************************************
DYNAMICS OPTION: Eulerian Mass Coordinate
alloc_space_field: domain 1, 86418536 bytes allocated
Input data is acceptable to use: fg
hybrid_opt = 2

use_theta_m = 1

WRF NUMBER OF TILES FROM OMP_GET_MAX_THREADS = 36
Tile Strategy is not specified. Assuming 1D-Y
....
...
Domain mapping info:
map_proj = 3
cen_lat = 0.225000E+02
cen_lon = 0.865000E+02
truelat1 = 0.225000E+02
truelat2 = 0.000000E+00
start_lat = 0.171538E+02
start_lon = 0.806196E+02
pole_lat = 0.900000E+02
dsm = 0.400000E+01

 

[JeremyB]

Hello priya_nit,

Sorry to hear it's not reading in. Unfortunately to see what's going on, we will need a few files. Can you please attach your: rsl.out.0000, rsl.error.0000, and namelist.input. If possible posting your observation file (ob.ascii) could help too. Without those, we can't properly diagnose the issue and provide a potential solution.

Thanks,
JeremyB

 

[Litta]

Hello JEremyB,
I also face similar issue.
I am trying to assimilate my surface observation (AWS data) using wrfda(v4.5.2). I produced the little.r format file using AWS data (F16-SYNOP) and ran the obsproc successfully. The data assimilation module worked successfully but the observation was not assimilated. I am trying to find the reason.

 

[JeremyB]

Hello Litta,

Unfortunately I'm not entirely sure what's going on. Everything looks good to me. Are you linking your 3DVAR/4DVAR files to ob.ascii (or ob01.ascii)? Or are you linking the obs file you provided me to the ob.ascii? That's the only thing I could think of.

Thanks,
JeremyB

 

[Litta]

Hello JeremyB

Thank you for your quick response. I used obs_gts_2024-02-11_00:00:00.3DVAR as ob.ascii.
I made 3DVAR run using test data provided by 3DVAR tutorial. It is working properly. But I didnit see any observation used list in rsl.out.0000
Please check attached file and image

 

[Litta]

Please check full data. The previous one is a sample test data

Thanks and best regards,
Litta

 

[Litta]

Hello JeremyB

I solved this issue. There was a mistake in namelist.input content. I wrote 2014 Instead of 2024

 

[JeremyB]

Hello Litta,

Glad you were able to fix it and thank you for sharing your solution.

Thanks,
JeremyB

 

[Litta]

Hello JeremyB,
Thank you for your support!
I did da_update_bc.exe after getting wrfvar_output. Then I copied wrfbdy_d01 and wrfvar_output wrfinput_d01. Then I ran wrf.exe.
But there is no difference in the wind plot at 00 UTC. Please check the image. Any suggestion?
Thanks
Litta

 

[JeremyB]

Hello Litta,

Can you please send your parame.in file? I wonder if there's something in there. I've seen this issue arise before and I'm not sure why it happens. Makes me think I should analyze my files as well in case this is happening to me. I suppose some other reasons could be your resolution? I know this isn't the best advice but I would mess around with it a bit. Maybe try having observations for time 1 and time 2? If you figure this out please let me know. I'll look into it on my side as well and report back if I find anything.

Thanks,
JeremyB

 

[JeremyB]

Hello Litta,

Something else is you might not have enough observations to make a difference. I just tried running WRFDA using prepbufr files (they contain quite a lot of observations), and I got a difference between the files. However, when I run using my little r files, I end up with identical files. I will try some more observations for my project and report back, but keep that in mind.

Edit: There's a way to determine if your observations were ingested into the model. Take a look at your rsl.out.0000 file. At the end it should tell you how many observations were ingested into the model. Here is an example where the mode ingested 47 observations. I'm learning with you so if you come across anything please share it.
Thanks,
JeremyB

 

[Litta]

Dear JeremyB,

Thank you for your support and suggestions!
I checked rsl.out.0000 files. 1139 observation datasets are ingested. But there is no difference in analysis data.
Please check my param.in and rsl.out.0000

 

[Litta]

Dear JeremyB,
Thank you for your suggestion. I never used diffwrf command. I tried now.
I saw some difference here (please check attached image). Why it is not showing in the image?
I used wrf-python for plotting. is that the issue of plotting?
Thanks and best regards,
Litta

 

[JeremyB]

Hello Litta,

The differences may be too slight to see, however with that number of observations I'm surprised it's so slight. WRF-python shouldn't impact it. You may want to ask an admin with this issue. I'm also curious to see what the results could be. If you do ask an admin, I would recommend attaching your namelist files as well as your rsl.out.0000 and rsl.error.0000 file as that will save some time.

Thanks,
JeremyB

 

[Litta]

Dear JeremyB,
Thank you for your suggestion.
Yes. I want to ask an admin this issue. I would like to know the reason.
Thanks and best regards,
Litta

 

[Litta]

Dear JeremyB,
Thank you for your suggestions.
I tried gdas.t00z.prepbufr.nr for data assimilation. It is worked well. But no difference in the output plot. Please check attachments. It is a different issue I think. Any suggestions.

with best regards,
Litta

 

[JeremyB]

Hello Litta,

I do see that you have some values in the first screenshot you sent, indicating there is a difference. When plotting, you likely won't see many differences because you have a larger domain and (assuming you use cartopy) the mapping can interpolate some values for contours, which can make it look similar. The best advice I can give to ensure that your files are different is to either run ncdiff (from the NCAR Command Language (ncl), this can be installed through conda or sudo apt install ncl if you're on ubuntu), or to create a python code (I used AI to generate a quick one and it worked perfectly) that can calculate the difference of the two files. I would then plot that file not in comparison to any other file. Differences will be very small, but you would then know if it worked. If you find that all the values are 0, then it didn't work properly. If neither of those worked, then like I suggested before, I would ask an admin. In my experience, I've tried assimilating with prepbufr files and that will typically work for me without an issue. I'm currently running into an issue where the prepbufr files aren't assimilating, however when it does assimilate, there is a difference in the file. If you haven't, I would also try WRFDA cycling as there you will see the differences in the wrfin files. I hope that helped and if you have any questions, feel free to ask.

Thanks,
JeremyB

 

[Litta]

Dear JeremyB

Thank you for your suggetion. I will check it.

Thanks,
Litta

 

[Litta]

Dear JeremyB,

I am not familiar with NCL.
I checked using python at 00 UTC. Eventhough figure didn't show much difference, there are some difference exists.
Please check the attachment.

Thanks and regards,
Litta

 

[JeremyB]

Dear JeremyB,

I am not familiar with NCL.
I checked using python at 00 UTC. Eventhough figure didn't show much difference, there are some difference exists.
Please check the attachment.

Thanks and regards,
Litta

Hello Litta,

NCL is very useful for performing quick analyses of wrf files. As for your plots, it looks like there are differences, so you should be good to go.

Thanks,
JeremyB