[Wrf-users] WRF with DS 090.0 (NNRP)

Wed Jan 10 17:26:11 MST 2007

Dear WRF Help and WRF users,

I'm trying to assimilate data DS090.0 (NNRP) that consist of surface files
(grb2d) and 3-d fileds (pgb) files. I use DS090.0 (NNRP) data and Vtable
NNRP for upper air data (pgb) and Vtable NNRPCFS for surface data (grb2d) in
wrfsi.nl (as described in manual - p.3-15) running wrfprep twice.
Applying siscan to hinterp.* I receive:

===============================================================
mti18 at l3n01-c /hpc/projects/WRF/wrfsi/domains/okura/siprd
:/hpc/projects/WRF/wrfsi/bin/siscan hinterp.d01.2002-12-04_12:00:00
Scanning hinterp.d01.2002-12-04_12:00:00  Domain Metadata Information
 -----------------------------------------------------
Domain Number .........  1
Parent ID .............  1
Dynamic Init. Source .. SI
Static Init. Source ... SI
Valid Date (YYYDDD) ... 2002338
Valid Time (sec UTC) .. 43200.0
Origin X in Parent ....     1
Origin Y in Parent ....     1
Nest Ratio to Parent ..     1
Delta X ............... 27000.0
Delta Y ............... 27000.0
Top Level .............   5000.0
Origin Z in Parent ....     0
X dimension ...........    30
Y dimension ...........    30
Z dimension ...........    14
  ---------------------------------------------------

 Variables found:
 NAME    S D   NX   NY   NZ  UNITS        DESCRIPTION       MINVAL   MAXVAL
AVGVAL
 ------- - - ----- ----- --- ------------ ---------------- -------- --------
--------
PRESSURE 0 1    14     0   0     Pa       Pressure levels     5000.  100000.
43536.
T        4 3    30    30  14     K        Temperature       208.060  293.402
241.207
U        4 3    30    30  14     m s-1    U                 -6.5135  46.2406
14.1638
V        4 3    30    30  14     m s-1    V                -20.2825   7.9488
-3.4748
HGT      4 3    30    30  14     m        Height               166.   20510.
8822.
RH       4 3    30    30  14              Relative Humidit   1.0000  87.5436
24.5830
PSFC     4 2    30    30   0     Pa       Surface Pressure   79427.  105562.
89615.
SM000010 1 2    30    30   0              Soil Moisture of  .000000  .333797
.173603
SM010200 1 2    30    30   0              Soil Moisture of  .000000  .251292
.112719
ST000010 1 2    30    30   0     K        T of 0-10 cm gro     .000  289.181
226.354
ST010200 1 2    30    30   0     K        T of 10-200 cm g     .000  294.372
233.757
SEAICE   1 2    30    30   0              Ice flag          .000000  .000000
.000000
SKINTEMP 1 2    30    30   0     K        Skin temperature  277.299  293.427
283.797
SNOW     1 2    30    30   0     kg m-2   Water equivalent  .000000  .000000
.000000
PMSL     4 2    30    30   0     Pa       Sea-level Pressu  101881.  103500.
102669.
SOILHGT  1 2    30    30   1     m        Terrain height o -9999.00  2008.15
-761.91
 End of file reached.
============================================================================
==========

So NNRP data have only 2 levels of soil temperature and moisture. That looks
consistent for me as NNRPSFC Vtable also has only 2 layers.
But in namelist.input for WRFV2 there is no option 2 layers:
(num_soil_layers                     = 2 <------------ 4/5/6 only)

Number of levels should be 5, 4 or 6. Of cause when I try to run real.exe
the model gives me a mistake (not enough levels):
==================================================
Time period #  61 to process = 2002-12-31_00:00:00.
Time period #  62 to process = 2002-12-31_12:00:00.
Total analysis times to input =   62.

----------------------------------------------------------------------------
-

 Current date being processed: 2002-12-01_00:00:00.0000, which is loop # 1
out of 62  configflags%julyr, %julday, %gmt: 2002 335 0.0000000000E+00
 -------------- FATAL CALLED ---------------
 FATAL CALLED FROM FILE:  module_initialize_real.b  LINE:     267
 No 10-40 cm, 0-10 cm, or 20 cm soil temperature data for TMN
 -------------------------------------------
====================================================  

When I try to remove sf_surface_physics (that is very bad) WRF gives me just
some rubbish. I didn't try NOAH and RUC LSM but I'm sure that the result
will be the same. Number of soil levels is equal to 5 by default.
So what configuration of namelist.input should I use to run WRF with NNRP
data?
Did anyone collide with this problem or did anyone manage to use DS090.0
(NNRP) data as input for ARF WRF 2.1.2 (or with WRF2.2).
"num_soil_layers" is still 4, 5 or 6 for WRF2.2. But if there are V-Tables
to use NNRP data there sould be an option (or trick) to run WRF with NNRP
data that I don't know.

Any suggestions are highly appreciable! 

Many thanks,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Mikhail Titov
Doctoral Candidate
Department of Geography
University of Canterbury - Te Whare Wananga o Waitaha Private Bag 4800
Christchurch New Zealand
Telephone:  +64 3 364 2987 ext 8839
Fax (Department): +64 3 364 2907
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

-----Original Message-----
From: WRF Help [mailto:wrfhelp at ucar.edu]
Sent: Saturday, 23 December 2006 3:16 p.m.
To: mti18 at student.canterbury.ac.nz
Subject: [SPAM: 3.000] Re: WRF with DS 090.0

Running grib_prep.pl is correct. But you should also put 3D data in a
separate directory than the 2D data. Your problem could result in grib_prep
not able to degrib certain data correctly. See
http://www.mmm.ucar.edu/wrf/users/docs/user_guide/
users_guide_appA.html
"Using Multiple Data Sources" for information too.

wrfhelp

On Dec 20, 2006, at 11:31 PM, mti18 at student.canterbury.ac.nz wrote: