Runtime Problems

Post Reply
jorge_p
Posts: 8
Joined: Tue Nov 20, 2018 3:45 pm

Runtime Problems

Post by jorge_p » Tue Nov 20, 2018 3:59 pm

Hi (I am a new WRFDA user),

I am trying to run 4DVAR for radar data (radial velocity and reflectivity). After succesfuly completing the Tutorial exercise for Radar assimilation (http://www2.mmm.ucar.edu/wrf/users/wrfd ... radar.html) I am running a test 4DVAR with my own radar data. WRFDA is version 3.8.1.

RUN-TIME Problems:

1. When I set the following options to "true"

use_radar_rv = .true.
use_radar_rf = .true.

The run aborts with this error (although cloud_cv_option is set to 1 in my namelist.input, see attached files namelist.input and rsl.error ).

---------------------------- FATAL ERROR -----------------------
Fatal error in file: <A HREF="http://www.mmm.ucar.edu/people/wrfhelp/ ... lve.inc</a> LINE: 216
Please set cloud_cv_options=1 for use_radar_rf
----------------------------------------------------------------
application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0

2. When I set the following options to "true" and "false"

use_radar_rv = .true.
use_radar_rf = .false.

The run is completed but after examining the output and the statistics, no data assimilation has been done (see attached files rsl.out and statistics ).


I have followed the steps described in the Tutorial presentation http://www2.mmm.ucar.edu/wrf/users/wrfd ... _Radar.pdf except for steps 2 and 4, thinking that this are not essential for a test run and can be tuned later) and I believe I have compiled WRFDA correctly by setting the mentioned environment variable (export CLOUD_CV=1) before running ./configure 4dvar.

My ob01.radar, ob02.radar etc. files are apparently without issues.

Also I am using the provided be.dat file in WRFDAV3/var/run/be.dat.cv3.

I would really appreciate any help on this...

Thank you in advance
Attachments
statistics.txt
(11.03 KiB) Downloaded 33 times
rsl.out.0000.txt
(14.98 KiB) Downloaded 46 times
rsl.error.0000.txt
(1.38 KiB) Downloaded 37 times
namelist.input
(4.19 KiB) Downloaded 39 times

kwthomas
Posts: 73
Joined: Tue Nov 06, 2018 10:25 pm

Re: Runtime Problems

Post by kwthomas » Tue Nov 20, 2018 8:26 pm

Hi...

I see this:

Cloud control variables are not implemented for cv_options = 3

My guess is this is your problem.
Kevin W. Thomas
Center for Analysis and Prediction of Storms
University of Oklahoma
Norman, Oklahoma

jorge_p
Posts: 8
Joined: Tue Nov 20, 2018 3:45 pm

Re: Runtime Problems

Post by jorge_p » Wed Nov 21, 2018 11:34 am

Hi Kevin,

thank you very much for your reply, yes I guess I underestimated that Warning message, I 'll try to generate my domain
specific be.dat and see the result for the
use_radar_rv = .true.
use_radar_rf = .true.

run.
Still I don't get why the
use_radar_rv = .true.
use_radar_rf = .false.
failed to run correctly which doesn't seem to depend on the cv_option = 3.

Thanks again

hclin2018
Posts: 31
Joined: Wed Jun 13, 2018 5:38 pm

Re: Runtime Problems

Post by hclin2018 » Tue Dec 04, 2018 1:03 am

cloud_cv_options is reset to be 0 in var/da/da_main/da_solve.inc when cv_options = 3.
The second line of the warning message is not printed due to a bug in the code.
The intended warning message should look like this:
--------------------------- WARNING ---------------------------
WARNING FROM FILE: <A HREF="http://www.mmm.ucar.edu/people/wrfhelp/ ... lve.inc</a> LINE: 205
Cloud control variables are not implemented for cv_options = 3
Resetting cloud_cv_options = 0 for cv_options = 3
---------------------------------------------------------------

In your rsl.out.0000.txt,
Observation summary
does not show observation counts.
Apparently the radar data are not read in properly.
You can set in namelist.input &wrfvar1 print_detail_radar=.true. to see if it gives you more information about reading your data.

The print should look like this if the radar data are read in successfully.
Observation summary
ob time 1
radar 20774 global, 7173 local
ob time 2
radar 24175 global, 7407 local
ob time 3
radar 12464 global, 5072 local

jorge_p
Posts: 8
Joined: Tue Nov 20, 2018 3:45 pm

Re: Runtime Problems

Post by jorge_p » Thu Dec 06, 2018 9:35 am

Thank you hclin2018,

I just saw your post, but yes I figured it out, it was a case of not reading the radar obs correctly, lat lon in the ob##.radar was written in reverse
by mistake and of course this meant that no obs where found inside my domain.

jorge_p
Posts: 8
Joined: Tue Nov 20, 2018 3:45 pm

Re: Runtime Problems

Post by jorge_p » Thu Dec 06, 2018 10:30 am

hclin2018 wrote:
Tue Dec 04, 2018 1:03 am
cloud_cv_options is reset to be 0 in var/da/da_main/da_solve.inc when cv_options = 3.
The second line of the warning message is not printed due to a bug in the code.
The intended warning message should look like this:
--------------------------- WARNING ---------------------------
WARNING FROM FILE: <A HREF="http://www.mmm.ucar.edu/people/wrfhelp/ ... lve.inc</a> LINE: 205
Cloud control variables are not implemented for cv_options = 3
Resetting cloud_cv_options = 0 for cv_options = 3
---------------------------------------------------------------

In your rsl.out.0000.txt,
Observation summary
does not show observation counts.
Apparently the radar data are not read in properly.
You can set in namelist.input &wrfvar1 print_detail_radar=.true. to see if it gives you more information about reading your data.

The print should look like this if the radar data are read in successfully.
Observation summary
ob time 1
radar 20774 global, 7173 local
ob time 2
radar 24175 global, 7407 local
ob time 3
radar 12464 global, 5072 local
Hi hclin2018, do you know the meaning of global and local ?

hclin2018
Posts: 31
Joined: Wed Jun 13, 2018 5:38 pm

Re: Runtime Problems

Post by hclin2018 » Thu Dec 06, 2018 8:19 pm

Global refers to the whole assimilation domain.
Local refers to the decomposed patch domain that each processor processes.

jorge_p
Posts: 8
Joined: Tue Nov 20, 2018 3:45 pm

Re: Runtime Problems

Post by jorge_p » Fri Dec 21, 2018 9:04 am

hclin2018 wrote:
Thu Dec 06, 2018 8:19 pm
Global refers to the whole assimilation domain.
Local refers to the decomposed patch domain that each processor processes.
Thank you again for your help

Post Reply

Return to “Radar DA”