WRF

Model description

Weather Research and Forecast model (WRF, Skamarock, 2008) is a limited area atmospheric model developed by a consortium of american institutions with contributions from the whole community. It is a non-hydrostatic primitive equations model used in a large variety of research areas.

Compilation

Model (version 3.9.1) has already been compiled in hydra

For compilation using the gcc copmilers see here WRFgcc

• From an instalation folder [INSTALLDIR] (/share/WRF, in hydra), creation of a folder structure

[INSTALLDIR]/[WRFversion]/[compiler]/[compilation_kind]/ 

• As example WRFv3.9.1 compiled with intel-compilers with distributed memory only

$ mkdir /share/WRF/WRFV3.9.1/ifort/dmpar/
$ cd /share/WRF/WRFV3.9.1/ifort/dmpar/

• Decompressing code source files

$ tar xvfz ../../../WRFV3.9.1.1.TAR.gz 
$ mv WRF WRFV3 
$ tar xvfz ../../../WPS.9.1.TAR.gz

• Compilation of model (wrf.exe, real.exe)
  • Going to the model source folder

cd WRFV3

• • Declaration of location of netcdf libraries and large file support (> 2GB)

$ export NETCDF='/usr/local'
$ export WRFIO_NCD_LARGE_FILE_SUPPORT=1

• • Running configuration of the compilation and picking up INTEL (ifort/icc) and distributed memory (dmpar) (option 15 and `basic' 1)

$ ./configure

• • Making a copy of the configuration file to make two modifications

$ cp configure.wrf configure.ifort.dmpar.wrf

• • Modifying the file configure.ifort.dmpar.wrf

#DM_FC          =      mpif90 -f90=$(SFC) 
#DM_CC          =      mpicc -cc=$(SCC) -DMPI2_SUPPORT
DM_FC          =      mpif90 
DM_CC          =      mpicc -DMPI2_SUPPORT

• • Using the modified file

$ cp configure.ifort.dmpar.wrf configure.wrf 
$ ./compile em_real >\& compile.log

• • If everything went fine one should have

$ ls main/*.exe
main/ndown.exe  main/real.exe  main/tc.exe  main/wrf.exe

• • And empty:

$ cat -n compile.log | grep Error
$ cat -n compile.log | grep undefined

• Compilation of the WRF-preprocessing (WPS: geogrid.exe, ungrib.exe, metgrid.exe)
  • Going there

cd ../WPS

• • Running configure and picking up 20 (Linux x86_64, Intel compiler (dmpar_NO_GRIB2))

configure

• • Compiling

./compile >& compile.log

• • If everything went right, one should have

$ ls */*.exe
geogrid/geogrid.exe  util/avg_tsfc.exe      util/int2nc.exe 
metgrid/metgrid.exe  util/calc_ecmwf_p.exe  util/mod_levs.exe
ungrib/g1print.exe   util/g1print.exe       util/rd_intermediate.exe
ungrib/g2print.exe   util/g2print.exe 
ungrib/ungrib.exe    util/height_ukmo.exe

Forcings

Atmospheric forcings

Provide the atmospheric conditions to the model at a given date.

• There is a shared space called /share
• At hydra all forcings are at:

/share/DATA/

• ERA-Interim
  • Thus, part of ERA-Interim forcings are:

/share/DATA/re-analysis/ERA-Interim/

• • Global monthly files at 0.75° horizontal resolution and all time-steps: 00, 06, 12, 18 are labelled as:
    • ERAI_pl[YYYY][MM]_[var1]-[var2].grib: pressure levels variables (all levels). GRIB codes as:
      • 129: geopotential
      • 157: relative humidty
      • 130: Temperature
      • 131: u-wind
      • 132: v-wind
    • ERAI_sfc[YYYY][MM].grib: all surface levels variables (step 0)
  • To download data:
    • Generate files from ECMWF web-page ERA-Interim
    • Go to the folder in hydra

cd /share/DATA/re-analysis/ERA-Interim/

• • • get the file (as link from ECMWF web-page (right bottom on `Download grib'), as e.g.:

$ wget https://stream.ecmwf.int/data/atls05/data/data02/scratch/_mars-atls05-a82bacafb5c306db76464bc7e824bb75-zn7P44.grib

• • • Rename file according to its content

$ mv _mars-atls05-a82bacafb5c306db76464bc7e824bb75-zn7P44.grib ERAI_sfc201302.grib

• ERA5
  • Thus, part of ERA5 forcings are:

/share/DATA/re-analysis/ERA5/

• 'NCEP-NNRP1'
  • Thus, full NCEP-NNRP1 forcings are (NOTE: no land data!!!):

/share/DATA/re-analysis/NCEP_NNRP/

Morphological forcings

Provide the geomorphological information for the domain of simulation: topography, land-use, vegetation-types, etc...

• In WRF there is a huge amount of data ans sources at different resolutions. At hydra everything is already there at:

/share/GEOG/

• They are ready to be use

Model use

WRF has two main parts:

• WPS: Generation of the domain, initial and boundary condition: Runs 4 programs:
  • geogrid: domain generation
  • ungrib: unpack atmospheric forcing \verb+grib+ files
  • metgrid: horizontal interpolation of the unpacked atmospheric forcing files at the domain of simulation
  • real: generation of the initial and boundary conditions using \verb+metgrid+ output
• WRF: model it self: wrf.exe

At hydra all the code is already compiled at (with folders WPS and WRFV3:

/share/WRF/[WRFversion]/[compiler]/[compilation_kind]/

As example WRFv3.9.1 compiled with intel-compilers with distributed memory only

/share/WRF/WRFV3.9.1/ifort/dmpar/

WPS

Let's assume that we work in a folder called WORKDIR (at users' ${HOME} at hydra). As example, let's create a two nested domain at 30 km for the entire South America and a second one at 4.285 km for Córdoba mountain ranges (see the WRF/namelist.wps)

WRF two domain configuration 30 and 4.285 km

• geogrid: generation of the domain
  • Creation of a folder for the geogrid section

$ mkdir geogrid
$ cd geogrid

• • Take the necessary files:

 $ ln -s /share/WRF/WRFV3.9.1/ifort/dmpar/WPS/geogrid/geogrid.exe ./
$ ln -s /share/WRF/WRFV3.9.1/ifort/dmpar/WPS/geogrid/GEOGRID.TBL.ARW  ./GEOGRID.TBL
$ cp /share/WRF/WRFV3.9.1/ifort/dmpar/WPS/geogrid/namelist.wps  ./

• • Domain configuration is done via namelist.wps (more information at: WPS user guide)

vim namelist.wps

• • Once it is defined, run it:

./geogrid.exe >& run_geogrid.log

• • It will create the domain files, one for each domain

geo_em.d[nn].nc

• • Some variables:
    • LANDMASK: sea-land mask
    • XLAT_M: latitude on mass point
    • XLONG_M: longitude on mass point
    • HGT_M: orographical height
  • At hydra the domain for SESA at 20 km for WRFsensSFC is located at

home/lluis.fita/estudios/dominios/CDXWRF

• ungrib: unpack grib files
  • Creation of the folder (from $WORKDIR)

$ mkdir ungrib 
$ cd ungrib

• • Linking necessary files from compiled source

$ ln -s /share/WRF/WRFV3.9.1/ifort/dmpar/WPS/ungrib/ungrib.exe ./
$ ln -s /share/WRF/WRFV3.9.1/ifort/dmpar/WPS/link_grib.csh ./

• • Creation of a folder for the necessary GRIB files and linking the necessary files (4 files per month)

$ mkdir GribDir
$ cd GribDir
$ ln -s /share/DATA/re-analysis/ERA-Interim/*201212*.grib ./ 
$ cd ..

• • Re-link files for WRF with its own script

./link_grib.csh GribDir/*

• • Should appear:

$ ls GRIBFILE.AA*
GRIBFILE.AAA\  GRIBFILE.AAB  GRIBFILE.AAC  GRIBFILE.AAD

• • We need to provide equivalences of the GRIB codes to the real variables. WRF comes with already defined GRIB equivalencies from different sources in folder Variable_Tables. In our case we use ECMWF ERA-Interim at pressure levels, thus we link

ln -s /share/WRF/WRFV3.9.1/ifort/dmpar/WPS/ungrib/Variable_Tables/Vtable.ERA-interim.pl ./Vtable

• • We need to take the domain file and get the right dates

$ cp ../geogrid/nameslist.wps ./

• • Files can be unpacked

./ungrib.exe >& run_ungrib.log

• • If everything went fine, should appear:

FILE:[YYYY]-[MM]-[DD]_[HH]

• And...

$ tail run_ungrib.log 
(...) 
********** 
Done deleting temporary files. 
********** 

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 
!  Successful completion of ungrib.   ! 
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

• metgrid horizontal interpolation of atmospheric forcing data at the domain of simulation
  • Creation of the folder (from $WORKDIR)

$ mkdir metgrid 
$ cd metgrid

• • Getting necessary files

$ ln -s /share/WRF/WRFV3.9.1/ifort/dmpar/WPS/metgrid/metgrid.exe ./ 
$ ln -s /share/WRF/WRFV3.9.1/ifort/dmpar/WPS/metgrid/METGRID.TBL.ARW ./METGRID.TBL

• • Getting the ungrib output

ln -s ../ungrib/FILE* ./

• • Link the domains of simulation

$ ln -s ../geogrid/geo_em.d* ./

• • Link the namelist from ungrib (to make sure we are using the same!)

$ ln -s ../ungrib/namelist.wps ./

• • Get the PBS (job queue script) to run the metrid.exe

$ cp /share/WRF/run_metgrid.pbs ./

• • And run it

qsub run_metgrid.pbs

• • If everything went fine one should have

met_em.d[nn].[YYYY]-[MM]-[DD]_[HH]:[MI]:[SS].nc

• • And...

$ tail run_metgrid.log
(...)
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 
!  Successful completion of metgrid.  ! 
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

• real vertical interpolation of atmospheric forcing at the domain of simulation
  • Creation of the folder (from $WORKDIR)

$ mkdir run 
$ cd run

• • Link all the necessary files from WRF (it already links all the necessary to run the model)

$ ln -s /share/WRF/WRFV3.9.1/ifort/dmpar/WRFV3/run/* ./

• • Remove and copy the configuration file (namelist.input)

$ rm namelist.input 
$ cp /share/WRF/WRFV3.9.1/ifort/dmpar/WRFV3/run/namelist.input ./

• • Edit the file and prepare the configuration for the run (re-adapt domain, physics, dates, output....). See an example for the two nested domain here WRF/namelist.input

$ vim namelist.input

• • Linking the metgrid generated files

ln -s ../metgrid/met_em.d*.nc ./

• • Getting the PBS job script for real.exe

$ cp /share/WRF/run_real.pbs ./

• • Getting a bash script to run executables in hydra (there is an issue with ulimit which does not allow to run with more than one node)

$ cp /share/WRF/launch_pbs.bash ./

• • And run it

qsub run_real.pbs

• • If everything went fine one should have (the basic ones):

wrfbdy_d[nn]  wrfinput_d[nn]  ...

• • • wrfbdy_d01: Boundary conditions file (only for the first domain)
    • wrfinput_d[nn]: Initial conditions file for each domain
    • wrffdda_d[nn]: Nudging file [optional]
    • wrflowinp_d[nn]: File with updating (every time-step of the atmospheric forcing) surface characteristics [optional]
  • And...

$ tail realout/[InitialDATE]-[EndDATE]/rsl.error.0000
(...)
real_em: SUCCESS COMPLETE REAL_EM INIT

WRF

• wrf Simulation (look on description of namelist for namelist configuration/specifications).
  • Getting the necessary PBS job (same folder for real)

$ cp /share/WRF/run_wrf.pbs ./

• • And run it

qsub run_wrf.pbs

• • If everything went fine one should have (the basic ones):

wrfout_d[nn]_[YYYY]-[MM]-[DD]_[HH]:[MI]:[SS]  wrfrst_d[nn]_[YYYY]-[MM]-[DD]_[HH]:[MI]:[SS]  ...

• • • wrfout/wrfout_d[nn]_[YYYY]-[MM]-[DD]_[HH]:[MI]:[SS]: simulation output (at η=(p-p_top)/(p_sfc-p_top) levels)
    • wrfout/wrfrst_d[nn]_[YYYY]-[MM]-[DD]_[HH]:[MI]:[SS]: restart file (to continue simulation)
    • wrfout/wrfxtrm_d[nn]_[YYYY]-[MM]-[DD]_[HH]:[MI]:[SS]: file with extremes from internal integration [optional]
    • wrfout/wrfpress_d[nn]_[YYYY]-[MM]-[DD]_[HH]:[MI]:[SS]: file at vertical pressure levels [optional]
    • namelist.output_[YYYY]-[MM]-[DD]_[HH]:[MI]:[SS]: All the parameters used for the simulation
    • stations: folder with the time-series files (tslist)
  • While running one can check the status on regarding on the rsl.error.0000 file, e.g.:

$ tail rsl.error.0000
(...)
Timing for main: time 2012-12-01_00:13:30 on domain  1:  3.86105 elapsed seconds

• • But if something went wrong:
    • CFL: At wrfout/[InitialDATE]-[EndDATE]/ there are the files rsl.[error/out].[nnnn] (two per cpu). Use to appear SIGSEV segmentation fault, and to look different sources, usually cfl. e.g. (usually look for the largest rsl file (after the 0000); [$ ls -rS rsl.error.*])

$ wrfout/20121201000000-20121210000000/rsl.error.0009
(...) 
d01 2012-12-01_01:30:00  33 points exceeded cfl=2 in domain d01 at time 2012-12-01_01:30:00 hours 
d01 2012-12-01_01:30:00  MAX AT i,j,k:    100    94    21 vert_cfl,w,d(eta)=    5.897325    17.00738    3.2057911E-02
(...)

Additional information

For more additional information and further details visit WRFextras

WRF4L: Lluís' WRF work-flow management

For information about Lluí's WRF work-flow management visit here WRF4L

CDXWRF: WRF for CORDEX

A new module developed in CIMA to attain CORDEX variable demands visit here CDXWRF