[Wrf-users] Problem with ARWpost.

Tue Oct 7 13:15:34 MDT 2008

Hello,

   I'm having a problem with the postprocessing. I am using the intel compilers, WRFV3 and it works perfectly. But when I try to postprocess it and try to see it with Grads , I get wrong data.

   Grads shows me all fields as constants (sometimes 0, sometimes 1000) but the first one, where I get what seems to be all the values of all the fields, all mixed up. 

----------------------------------------------------------------
ga-> set t 1 24
Time values set: 2003:1:1:0 2003:1:1:23 
ga-> d pressure
Notice:  Automatic Grid Interpolation Taking Place
Constant field.  Value = 1000
Contouring: 0.1 to 0.116 interval 0.002 
Too many contour levels -- adjusting cint
Contouring: 295.8 to 302 interval 0.2 
Contouring: 22.6 to 28.8 interval 0.2 
Contouring: 1000 to 1000 interval 0.2 
Contouring: 0.2 to 0 interval 0.2 
Contouring: 295 to 301.2 interval 0.2 
Contouring: 21.8 to 28 interval 0.2 
Contouring: 1000 to 1000 interval 0.2 
Contouring: 0.2 to 0 interval 0.2 
Contouring: 294.4 to 299.6 interval 0.2 
Contouring: 21.2 to 26.4 interval 0.2 
Contouring: 1000 to 1000 interval 0.2 
Contouring: 0.2 to 0 interval 0.2 
-------------------------------------------------------------------

Relevant parts of my namelist.ARWpost are:

&datetime
 start_date = '2003-01-01_00:00:00',
 end_date   = '2003-01-01_23:00:00',
 interval_seconds = 3600,
 tacc = 0,
 debug_level = 1,
/

&io
 io_form_input  = 2,
 input_root_name = '/datos/WRF3/WRFV3/run/wrfout_d03_2003-01-01_00:00:00'
 output_root_name = './test'
 plot = 'list'
 fields = 'height,pressure,tk,tc'
 output_type = 'grads'
 mercator_defs = .true.
/

&interp
 interp_method = 1,
 interp_levels = 1000.,950.,900.,850.,800.,750.,700.,650.,600.,550.,500.,450.,400.,350.,300.,250.,200.,150.,100.,
/

We think that WRF's output is ok, because we could postprocess the same files in another, older cluster, with WRF2GrADS. Also, we used ncdump on the WRF output and the data looks correct.

We are using the intel compilers v. 10.1 , WRF 3 , netcdf 4.0 , in a 2-nodes cluster (16 procs), Fedora 8, 64 bits. 

How can I debug this? Any help will be appreciated. 

Thanks in advance,

 -- Moira

«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»
Moira E. Doyle
Dto. Ciencias de la Atmósfera y los Océanos
Facultad de Ciencias Exactas y Naturales. UBA.
CIMA - CONICET
Int. Güiraldes 2620. Pabellón II. 2do Piso. Ciudad Universitaria
1428. Ciudad de Buenos Aires. Argentina
Tel: (54 11) 4576 3356/3364 ext. 28
Fax: (54 11) 4576 3356/3364 ext. 12

doyle at cima.fcen.uba.ar
medoyle_ar at yahoo.com
«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»

-- 
Este mensaje ha sido analizado por el Servidor de Mail de CIMA
en busca de virus y otros contenidos peligrosos,
y se considera que esta limpio.

-------------- next part --------------
Hello,

   I'm having a problem with the postprocessing. I am using the intel compilers, WRFV3 and it works perfectly. But when I try to postprocess it and try to see it with Grads , I get wrong data.

   Grads shows me all fields as constants (sometimes 0, sometimes 1000) but the first one, where I get what seems to be all the values of all the fields, all mixed up.

----------------------------------------------------------------
ga-> set t 1 24
Time values set: 2003:1:1:0 2003:1:1:23
ga-> d pressure
Notice:  Automatic Grid Interpolation Taking Place
Constant field.  Value = 1000
Contouring: 0.1 to 0.116 interval 0.002
Too many contour levels -- adjusting cint
Contouring: 295.8 to 302 interval 0.2
Contouring: 22.6 to 28.8 interval 0.2
Contouring: 1000 to 1000 interval 0.2
Contouring: 0.2 to 0 interval 0.2
Contouring: 295 to 301.2 interval 0.2
Contouring: 21.8 to 28 interval 0.2
Contouring: 1000 to 1000 interval 0.2
Contouring: 0.2 to 0 interval 0.2
Contouring: 294.4 to 299.6 interval 0.2
Contouring: 21.2 to 26.4 interval 0.2
Contouring: 1000 to 1000 interval 0.2
Contouring: 0.2 to 0 interval 0.2
-------------------------------------------------------------------

Relevant parts of my namelist.ARWpost are:

&datetime
 start_date = '2003-01-01_00:00:00',
 end_date   = '2003-01-01_23:00:00',
 interval_seconds = 3600,
 tacc = 0,
 debug_level = 1,
/

&io
 io_form_input  = 2,
 input_root_name = '/datos/WRF3/WRFV3/run/wrfout_d03_2003-01-01_00:00:00'
 output_root_name = './test'
 plot = 'list'
 fields = 'height,pressure,tk,tc'
 output_type = 'grads'
 mercator_defs = .true.
/
&interp
 interp_method = 1,
 interp_levels = 1000.,950.,900.,850.,800.,750.,700.,650.,600.,550.,500.,450.,400.,350.,300.,250.,200.,150.,100.,
/
We think that WRF's output is ok, because we could postprocess the same files in another, older cluster, with WRF2GrADS. Also, we used ncdump on the WRF output and the data looks correct.

We are using the intel compilers v. 10.1 , WRF 3 , netcdf 4.0 , in a 2-nodes cluster (16 procs), Fedora 8, 64 bits.

How can I debug this? Any help will be appreciated.

Thanks in advance,

 -- Moira

«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»
Moira E. Doyle
Dto. Ciencias de la Atmósfera y los Océanos
Facultad de Ciencias Exactas y Naturales. UBA.
CIMA - CONICET
Int. Güiraldes 2620. Pabellón II. 2do Piso. Ciudad Universitaria
1428. Ciudad de Buenos Aires. Argentina
Tel: (54 11) 4576 3356/3364 ext. 28
Fax: (54 11) 4576 3356/3364 ext. 12

mailto:doyle at cima.fcen.uba.ar doyle at cima.fcen.uba.ar
mailto:medoyle_ar at yahoo.com medoyle_ar at yahoo.com
«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»«»

-- 
Este mensaje ha sido analizado por el Servidor de Mail de CIMA
en busca de virus y otros contenidos peligrosos,
y se considera que esta limpio.