Weather Research and Forecasting

(WRF) Modeling System

 What is WRF?
• WRF: Weather Research and Forecasting
Model
– Now with ARW and NMM dynamic
components

• Its development is led by NCAR/MMM,

NOAA/GSD and NOAA/NCEP/EMC with
partnerships at AFWA, FAA, NRL and
collaborations with universities and other
government agencies

• Includes research and operational models

 What can WRF be used for?
• Idealized simulations at many scales (e.g.
convection, baroclinic waves, large eddy
simulations)
• Atmospheric physics/parameterization
research
• Data assimilation research
• Case-study research
• Real-time NWP and forecast system research
• Regional climate and seasonal time-scale
research
• Coupled-model (e.g. ocean, chemistry)
applications
• Teaching
 Who uses WRF?
• Academic atmospheric scientists
• Forecast teams at operational
centers
• Air Quality scientists
• Others
 WRF modeling system flowchart

WRF
Terrestial data
RIP4

WPP
WRF Model (GrADS,
WPS Post GEMPAK)
ARW/NMM processing

Real data: NCL

NAM, GFS,
RUC, NRRP
ARWPost
AGRMET (Soil)
(GrADS,
Vis 5D)
 Download and Building WRF code
 Get the WRF zipped tar file of latest WRF and WPS from

http://www.mmm.ucar.edu/wrf/users/download/get_source.html

How to access the University of Florida Server

 ssh –X username@hiypergator.rc.ufl.edu
password: (type ur password)
You can also download the Meteorological data from the below link

http://dss.ucar.edu/dsszone/ds083.2/

 WRF should be install first before WPS

You can download the WRF-ARW user’s guide from the link
http://www.mmm.ucar.edu/wrf/users/docs/user_guide_V3/contents.html
 Unzip and untar the downloaded tar file type the following command

$gunzip WRFV3.1.1.TAR.gz and then

$tar –xvf WRFV3.1.1.TAR (It will create a directory WRFV3/ )
$cd WRFV3 ( you will see the following top level directories)

arch/ configure* dyn_exp/ frame/ inc/ Makefile

share/ clean* dyn_nmm/ inc/ hys/
Registry test/ compile* main/ dyn_em/
external/ run/ tools/ README README_test_cases
README.rsl_output

 Then configure and compile the WRF model

 Before configure and compile the WRF model you have to set the paths of
compilers and libraries inside .bashrc file
 vi .bashrc (it will open the bashrc file and edit accordingly and save it)
 Set PATH for required bin and lib directories

PATH=/usr/bin:/opt/NETCDF/intel_NETCDF/bin:…:$PATH
(here you have to set the path to bin files of all libraries and compilers)
LD_LD_LIBRARY_PATH=/lib:/usr/lib:/usr/local/lib:/opt/NETCDF/
intel_NETCDF/lib:$LD_LIBRARY_PATH (here you have to set the path to
library files of all libraries and compilers)
INCLUDE=/usr/include:/opt/mpich2_intel/include:/usr/local/include:/opt/
NETCDF/intel_NETCDF/include::/opt/intel/fce/10.0.026/lib:$INCLUDE (here you
have to set the path to include files of all libraries and compilers)

export WRF_EM_CORE=1
export WRF_NMM_CORE=0
Export NETCDF=/opt/NETCDF/intel_NETCDF (the most important library
required to run WRF model)
 ./configure (It will ask to enter one from the following option)

Please select from among the following supported platforms.

1. Linux x86_64, PGI compiler with gcc (serial)

2. Linux x86_64, PGI compiler with gcc (smpar)
3. Linux x86_64, PGI compiler with gcc (dmpar)
4. Linux x86_64, PGI compiler with gcc (dm+sm)
5. Linux x86_64 i486 i586 i686, ifort compiler with icc (serial)
6. Linux x86_64 i486 i586 i686, ifort compiler with icc (smpar)
7. Linux x86_64 i486 i586 i686, ifort compiler with icc (dmpar)
8. Linux x86_64 i486 i586 i686, ifort compiler with icc (dm+sm)
9. Linux i486 i586 i686 x86_64, PathScale compiler with pathcc (serial)
10. Linux i486 i586 i686 x86_64, PathScale compiler with pathcc (dmpar)
11. x86_64 Linux, gfortran compiler with gcc (serial)
12. x86_64 Linux, gfortran compiler with gcc (smpar)
13. x86_64 Linux, gfortran compiler with gcc (dmpar)
14. x86_64 Linux, gfortran compiler with gcc (dm+sm)

Compile for nesting? (0=no nesting, 1=basic, 2=preset moves, 3=vortex

following) [default 0]: 0 (choose 0)
 Compile WRF for real case
$./compile em_real >& compile.log &
(To see the background process type the command
$ tail –f compile.log )
$ ls -ls main/*.exe (check this command after compilation)
 if you are compiling for a real-data case, you should see
ndown.exe, real.exe, and wrf.exe inside the main/ directory
after the successful compilation.

Building WPS Code

 gunzip WPSV3.1.1.TAR.gz
 tar –xvf WPSV3.1.1.TAR (It will untar and create a directory WPS/ )
 cd WPS
 Configure and compile WPS
 $ ./configure
Please select one among the following supported platforms.

1. PC Linux x86_64, Intel compiler serial, NO GRIB2

2. PC Linux x86_64, Intel compiler serial
3. PC Linux x86_64, Intel compiler DM parallel, NO GRIB2
4. PC Linux x86_64, Intel compiler DM parallel
5. PC Linux x86_64 (IA64 and Opteron), PGI compiler 5.2 or higher, serial, NO
GRIB2
6. PC Linux x86_64 (IA64 and Opteron), PGI compiler 5.2 or higher, serial
7. Cray XT Linux x86_64 (IA64 and Opteron), PGI compiler 5.2 or higher, DM
parallel, NO GRIB2
8. PC Linux x86_64 (IA64 and Opteron), PGI compiler 5.2 or higher, DM parallel,
NO GRIB2
9. PC Linux x86_64 (IA64 and Opteron), PGI compiler 5.2 or higher, DM parallel
10. PC Linux x86_64 (IA64 and Opteron), PathScale compiler 2.1 or higher, serial,
NO GRIB2
11. PC Linux x86_64 (IA64 and Opteron), PathScale compiler 2.1 or higher, DM
parallel, NO GRIB2
 $ ./compile >& compile.log &
$ ls -l *.exe (type this command after compilation)
you should see geogrid.exe, ungrib.exe, and metgrid.exe files for a successful
compilation
Different components of WPS
Geogrid
The purpose of geogrid is to define the simulation
domains, and interpolate various terrestrial data sets to the
model grids.
Ungrib
 The Ungrib program reads GRIB file and extract
interpolated metgrid output
meteorological fields from GRIB file. (The GRIB files contain
can then be
time-varying meteorological ingested
fields and are typically from
by
another global the real.exe program.
model)
Metgrid
The metgrid program horizontally interpolates the
meteorological data that are extracted by the ungrib program
onto the model grid defined by the geogrid program. The
interpolated metgrid output can then be ingested by the
real.exe program.
 namelist.wps: parameters setting for WRF
preprocessor (WPS)
&share
wrf_core = 'ARW',
max_dom = 1, (for single domain: only edit 1st column)
start_date = '2011-11-09_00:00:00','2006-08-16_12:00:00', Change the dates
according your case
end_date = '2011-11-10_00:00:00','2006-08-16_12:00:00',
interval_seconds = 21600 (Boundary conditions supplied time interval (sec)

to the model)
io_form_geogrid = 2,
/
&geogrid
parent_id = 1, 1,
parent_grid_ratio = 1, 3,
i_parent_start = 1, 31,
j_parent_start = 1, 17,
e_we = 130, 112, (No. of grid points in WE direction)
e_sn = 120, 97, (No. of grid points in SN direction)
geog_data_res = ‘10m','2m', (geographical data resolution)
dx = 27000, (Horizontal grid resolution (m) in WE direction)
dy = 27000, (Horizontal grid resolution (m) in SN direction)
 Continue…

map_proj = 'mercator', Horizontal map-coordinate

ref_lat = 16.4, Central latitude options
ref_lon = 81.3, Central longitude
truelat1 = 16.4, 1. Lambert Conformal:
truelat2 = 0.0,
stand_lon = 81.3, Mid-latitude applications
geog_data_path = '/usr/geog‘ (Path of geographical data set)
2. Polar Stereographic:
/ High-latitude applications
&ungrib 1. Mercator:
out_format = 'WPS',
prefix = 'FILE', Low-latitude applications
/

&metgrid
fg_name = 'FILE'
io_form_metgrid = 2,
 WRF ARW Initialization for Real Data
/WPS

geogrid.exe geo_em.d01.nc WPS output

(met_em.d01)

Setup the Vtable for GFS data set

$ln -s ungrib/Variable_Tables/Vtable.GFS Vtable /WRFV3/
test/em_real

Setup the path for the meteorological data

real.exe
$./link_grib.csh
/scratch/lfs/nyrojasr/WRF/Data/fnl*
ungrib.exe metgrid.exe
FILE:yyyy-mm-dd_hh met_em.d01.yyyy-mm-dd_hh:mm:ss.nc Wrfinput Wrfbdy
_d01 _d01

wrf.exe
wrfout_d01_yyyy-mm-dd_hh:mm:ss
namelist.input: parameters setting for WRF model run
Example for a single domain run: max_dom=1
&time_control
run_days = 0,
run_hours = 24,
run_minutes = 0,
run_seconds = 0,
start_year = 2011, 2010, 2010,
start_month = 11, 01, 01,
start_day = 09, 24, 24,
start_hour = 00, 00, 00,
start_minute = 00, 00, 00,
start_second = 00, 00, 00,
end_year = 2011, 2010, 2010,
end_month = 11, 01, 01,
end_day = 10, 25, 25,
end_hour = 00, 00, 00,
end_minute = 00, 00, 00,
end_second = 00, 00, 00,
interval_seconds = 21600
input_from_file = .true.,.true.,.true.,
history_interval = 180, 180, 60,
frames_per_outfile = 1000, 1000, 1000,
restart = .false.,
restart_interval = 5000,
io_form_history = 2
Example for a nested run: max_dom=1
&domains
time_step = 162, (6*dx(km))sec
time_step_fract_num = 0,
time_step_fract_den = 1,
max_dom = 1,
s_we = 1, 1, 1,
e_we = 130, 112, 94,
s_sn = 1, 1, 1,
e_sn = 120, 97, 91,
s_vert = 1, 1, 1,
e_vert = 28, 28, 28,
dx = 27000, 10000, 3333,
dy = 27000, 10000, 3333,
grid_id = 1, 2, 3,
parent_id = 0, 1, 2,
i_parent_start = 0, 31, 30,
j_parent_start = 0, 17, 30,
parent_grid_ratio = 1, 3, 3,
parent_time_step_ratio = 1, 3, 3,
feedback = 1,
smooth_option = 0
&physics
mp_physics = 4, 5, 5, 5, Micro_physics
ra_lw_physics = 1, 1, 1, 1,
ra_sw_physics = 1, 1, 1, 1,
radt = 10, 10, 10, 10,
sf_sfclay_physics = 5, 1, 1, 1,
sf_surface_physics = 2, 1, 1, 1,
bl_pbl_physics = 5, 1, 1, 1,
bldt = 0, 0, 0, 0,
cu_physics = 3, 3, 0, 0, Cumulus scheme
cudt = 5, 5, 5, 5,
isfflx = 1,
ifsnow = 0,
icloud = 1,
surface_input_source = 1,
num_soil_layers = 4,
sf_urban_physics = 0,
maxiens = 1,
maxens = 3,
maxens2 = 3,
maxens3 = 16,
ensdim = 144,
 WRF Parameterizations
Microphysics (mp_physics)
Kessler scheme: This is used for idealized cloud modeling studies. Warm-rain only
no ice (mp_physics=1).
Lin et al. scheme: A sophisticated scheme that has ice, snow and graupel processes,
suitable for real-data high-resolution simulations (2).
WSM 3-class scheme: A simple efficient scheme with ice and snow processes suitable
for mesoscale grid sizes (3).
WSM-5 class scheme: A slightly more sophisticated version of WSM-3 that allows
for mixed-phase processes and super-cooled water (mp_physics=4).
Ferrier (New Eta) scheme: The operational microphysics in NCEP models. A simple
efficient scheme with diagnostic mixed-phase processes (5).
WSM 6-class scheme: A scheme with ice, snow and graupel processes suitable for
high-resolution simulations (6).
 Goddard microphysics scheme: A scheme with ice, snow and graupel processes
suitable for high-resolution simulations (7).
Thompson et al. graupel scheme: 6-class microphysics with graupel, ice number
concentration also predicted (8).
Cumulus scheme (cu_physics):

Kain-Fritsch scheme: Deep and shallow convection sub-grid scheme

using a mass flux approach with downdrafts and CAPE removal time
scale (cu_physics = 1).

Betts-Miller-Janjic scheme: Operational Eta scheme. Column moist

adjustment scheme relaxing towards a well-mixed profile (2).

Grell-Devenyi ensemble scheme: Multiple-parameter, explicit

updrafts/downdrafts (cu_physics = 3).
 WRF-Postprocessor-ARWpost
 ARWpost is the module used in WRF for post-processing the model output
-- Necessary software required- GrADS

First You have to install GrADS for running ARWpost module

 unzip and untar the tar file located at

/tmp/WRF
 cd ARWpost
 ./configure
Please select from among the following supported platforms.

1. PC Linux i486 i586 i686 x86_64, PGI compiler (no vis5d)

2. PC Linux i486 i586 i686 x86_64, PGI compiler (vis5d)
3. PC Linux i486 i586 i686 x86_64, Intel compiler (no vis5d)
4. PC Linux i486 i586 i686 x86_64, Intel compiler (vis5d)
5. PC Linux i486 i586 i686 x86_64, gfortran compiler (no vis5d)
6. PC Linux i486 i586 i686 x86_64, gfortran compiler (vis5d)
Enter selection [1-6] : 3
  After configuration check the configure.arwp file and edit the path for WRFV3
 ./compile
After successful compilation it will create the execution file under /src directory
/src/ARWpost.exe
 Afer that you have to edit the namelist.ARWpost file for running ARWpost

 $vi namelist.ARWpost
&datetime
start_date = '2011-11-09_00:00:00',
end_date = '2011-11-10_00:00:00',
interval_seconds = 3600,
tacc = 180,
debug_level = 0,
/
&io
io_form_input = 2,
input_root_name = './wrfout_d01_2011-11-09_00:00:00‘ (path to input file)
output_root_name = './Expt-2011‘ (output file name)
plot = 'all_list'
fields = 'height,pressure,tk,tc,theta,td,td2,rh,rh2,umet,vmet,pressure,u10m,v10m,
wdir,wspd,wd10,ws10,slp,mcape,mcin,lcl,lfc,cape,cin,dbz,max_dbz,clfr'
output_type = 'grads'
mercator_defs = .true.
/
split_output = .true.
frames_per_outfile = 2

output_type = 'grads'
output_type = 'v5d'

plot = 'all'
plot = 'list'
plot = 'all_list'
! Below is a list of all available diagnostics
fields = 'height,geopt,theta,tc,tk,td,td2,rh,rh2,umet,vmet,pressure,u10m,v10m,
wdir,wspd, wd10,ws10,slp,mcape,mcin,lcl,lfc,cape,cin,dbz,max_dbz,clfr'
&interp
interp_method = 1,
interp_levels = 1000.,990.,980.,970.,960.,950.,900.,850.,800.,750.,700.,
650.,600.,550.,500.,450.,400.,350.,300.,250.,200.,150.,100.,
/
extrapolate = .true.

interp_method = 0, ! 0 is model levels, -1 is nice height levels, 1 is user

specified pressure/height

interp_levels =
1000.,950.,900.,850.,800.,750.,700.,650.,600.,550.,500.,450.,400.,350.,300.,25
0.,200.,150.,100.,
interp_levels = 0.25, 0.50, 0.75, 1.00, 2.00, 3.00, 4.00, 5.00, 6.00, 7.00, 8.00,
9.00, 10.0, 11.0, 12.0, 13.0, 14.0, 15.0, 16.0, 17.0, 18.0, 19.0, 20.0,

 After successful edit the namelist.ARWpost you have to run

./ARWpost.exe
 It will create the following files
output_file_name.ctl
output_file_name.dat
 Now you can visualize the output file using GrADS software
 Type the following commands
$ grads (It will open the graphics window)
$ open output_file_name.ctl (it will open the ctl file)
$ q file (It will show the list of variables in the file)
$ set t 3 (Set the time to display the variable at that time)
$ d (variable name) (choose the variable to display)
ga-> open 20100820-d3-CTRL3hrly-ThGD.ctl
Output file (*.ctl)
Scanning description file: 20100820-d3-CTRL3hrly-ThGD.ctl
Data file 20100820-d3-CTRL3hrly-ThGD.dat is open as file 1
LON set to 75.5019 78.3419
LAT set to 27.1023 29.5257
LEV set to 1000 1000
Time values set: 2010:8:20:0 2010:8:20:0
E set to 1 1 Station/city Latitude
ga-> set lat 28.5 and longitude (Delhi)
LAT set to 28.512 28.512
ga-> set lon 77.2
LON set to 77.2117 77.2117
ga-> set t 1 9
Time values set: 2010:8:20:0 2010:8:21:0
ga-> d t2-273.15
ga-> draw ylab Temperature (deg C)
ga-> draw xlab Time (UTC)
ga-> draw title 24hr-temperature2m-Delhi
ga->
Command to extract ASCII data to compare with observation data

ga-> set gxout print

ga-> set prnopts %g 1
ga-> set t 1 9 (set time from 1 to 9 )
time set from 00UTC 20th Aug to
Time values set: 2010:8:20:0 2010:8:21:0
00UTC 21st Aug 2010
ga-> d t2-273.15
Printing Grid -- 9 Values -- Undef = -9.99e+08
26.0597 (00UTC 20 Aug)
27.3945 (03UTC)
30.3203 (06UTC)
31.9206 (09UTC)
30.2627 (12UTC)
28.1149 (15UTC)
26.683 (18UTC)
25.9813 (21UTC)
25.4863 (00UTC 21 Aug)
ga->
WRF-ARW Simulation of LAILA Cyclone
(17-21 May, 2010)

Observed
Track
Cyclone – LAILA (17-21 May 2010)

Central Lat= 16.4 N

Central Lon= 81.3 E

No. of grid points:

WE-direction= 130
SN-direction = 120

Horizontal resolution:
Dx= 27km
Dy=27km
Model simulated Wind field from 19 May-20 May 2010
Model simulated Rainfall (mm) from 19 May-20 May 2010
Model simulated Sea Level Pressure (hPa) from 19 May-20 May 2010
Model simulated Rainfall (mm) from 19 May-20 May 2010