Calmet run error

141 views
Skip to first unread message

Abdullah Butt

unread,
Mar 19, 2023, 10:34:34 AM3/19/23
to european-calpuff-user-discussion-group
Hi, 

I am running calmet for a 3 day run output file from Wrf and I am encountering the following errors repeatedly. 

1. Meteo grid is not fully inside of 3D.dat file 
2. Count of layers in the prognostic wind model's grid (99) exceeds the limit (MXNZP=50)

What layers is calmet referring to here? I am using the ds083.2 datasets. There is no information on layers. Where can I look for this detail in wrf or metgrid? Previously when I ran wrf for a smaller domain size the layers error said it was 55 layers now with a bigger domain size to accommodate the first error the layers now increased to 99 and the both errors are still there. 

Please guide me if anyone has any knowledge pertaining to these errors. 

Thanks!

mert BAYRAKDAR

unread,
Mar 19, 2023, 12:17:36 PM3/19/23
to european-calpuff-user-discussion-group
Hi Abdullah, 

First, my apologize for my English knowledge.
1. WRF data needs to be inside of the grid domain, for example, my domain grid size 31 x 31,  3000 m. I can work under these limited areas. I'm generally using ds083.3 data but it doesn't matter your situation.  You can use "ncdump -h wrf_file", it gives you much more information about your data. If you want to resize of your WRF data, for example, a 20 x 20 250m cartesian grid. You need to use calwrf model. Calwrf model will be convert to "WRF.out" data to calmet ".dat" file. Then you can resize what you want to use area size or you can just use depend on wrf domain size area.
2. I'dont know what is this error.

Best regard.

19 Mart 2023 Pazar tarihinde saat 17:34:34 UTC+3 itibarıyla abdullah...@gmail.com şunları yazdı:

Abdullah Butt

unread,
Mar 19, 2023, 2:32:15 PM3/19/23
to european-calpuff-user-discussion-group
Hi, 

Thank you for suggesting a solution to the first problem. I will definitely work as advised. 

As far as second error is concerned I really am puzzled as I can't find anything online even. Will however post here if I find a solution so if anybody else encounters it they might get some help. 

Thanks! 

Abdullah Butt

unread,
Apr 1, 2023, 7:29:39 AM4/1/23
to european-calpuff-user-discussion-group
Update 

The error on layers has been resolved. Turns out there was some incompatibility with the CALWRF version I was using to generate the 3D.dat file. The error went away when I used a different CALWRF version.

mert BAYRAKDAR

unread,
Apr 1, 2023, 3:52:42 PM4/1/23
to european-calpuff-user-discussion-group
Hi abdullah, could you share the calwrf error and which version was used for calwrf ? 
1 Nisan 2023 Cumartesi tarihinde saat 14:29:39 UTC+3 itibarıyla abdullah...@gmail.com şunları yazdı:

Abdullah Butt

unread,
Apr 1, 2023, 5:02:27 PM4/1/23
to european-calpuff-user-discussion-group
Sure. First thing Monday! I work at a lab and it's at work PC. 

Gianfranco De troia

unread,
Aug 27, 2023, 2:25:59 PM8/27/23
to european-calpuff-user-discussion-group
Hi Mertbayr
I tried to use  ds083.3 without use it as WRF input. I only converted it in  netCDF format by means of  ncl_converter2nc. After  setup .inp file setup I execute calwrf , and the programm stops showing this message:

C:\Users\Gianfranco\Desktop\binary_windows>C:\Users\Gianfranco\Desktop\binary_windows\calwrf.exe C:\Users\Gianfranco\Desktop\binary_windows\calwrf.inp
  Control inp file:C:\Users\Gianfranco\Desktop\binary_windows\calwrf.inp
  2D.DAT flag input not exist, set to one
  Set 2D.DAT flag to 1:           1
  Default 2D.DAT filename:test1.m2d
  Open WRF netcdf file            1 : gdas1.fnl0p25.2015070800.f00.nc

 Processing GLOBAL ATTRIBUTES:

 This is not a wrfout file
 No 3D.DAT will be created

I need to manipulate ds083.3 data file in such a way?
Thanks

anel20...@gmail.com

unread,
Jun 1, 2024, 7:44:26 AM6/1/24
to european-calpuff-user-discussion-group
Hi ds083.3,

calwrfconverts WRF output files to calmet input files. To run calwrf you need real WRF output files.

you can use ds083.3to input WRF. not to calwrf.

good luck,
A
Reply all
Reply to author
Forward
0 new messages