How to Compile?

How to Run?

WRF Namelist

Known Problems and Fixes

Graphic Tools





WRF Model Version 3.4.1: UPDATES

WRF Model Update

WRF model tar file has been updated to Version 3.4.1 on Aug 16, 2012. For known problems in V3.4.1, please click here.

WRF Pre-Processing System (WPS) has been updated to Version 3.4.1.

Note: The V3.4.1 WRF can use wrfinput and wrfbdy files from earlier V3 versions only if one sets hypsometric_opt = 1 in namelist &domains. However if you would like to use newer options introduced in 3.4, you may benefit from the change if wrfinput and wrfbdy files are regenerated.

As always, be cautious when using new options.

If you are interested in seeing how V3 has been tested, click here.

Improvements and Bug fixes: (with the significant ones highlighted in bold font)


- Resetting cudt = 0 for BMJ, all SAS schemees, and Tiedtke since the time-step rain isn't correctly calculated when adaptive time is used. The time-step rain is used in LSM.
- Bug fix for cumulus calls if adaptive time stepping is used with OpenMP, either by setting numtiles to > 1 or OMP_NUM_THREADS > 1. The bug caused cumulus calls to only happen in the first tile (Thanks to Todd Hutchinson of WSI).
- Glacier ice treatment has been moved out Noah code, and replaced by a call from surface driver to seaice_noah.
- NoahMP: bug fixes and its namelists have been changed to single value or non-domain-dependent.
- MYNN sfc layer: Add iz0tlnd option, modify isftcflx options (see README.namelist for details)
- MYNN PBL: add option to output diagnostics (bl_mynn_tkebudget = 1), some tuning of the scheme
- YSU: fix for stable surface conditions (Thanks to Heather Richardson and Sukanta Basu at North Carolina State University) and consistency with thermal roughness length. The change may result in improvement in surface wind forecast at night.
- YSU: a bug fix for nest starting at later time due to introduction of 'topo_wind' option, even if the option isn't used.
- QNSE/EDMF: bug fix for EDMF (day-time) (Thanks to Steven Decker of Rutgers)
- SSiB: activate seaice function
- isftcflx = 2: Cd and Ck changed to formula used in 2012 ARW RT runs (not yet in revised surface scheme, sf_sfclay_physics = 11)
- Thompson MP: update to graupel intercept parameter and others
- GD and G3 schemes: correct the bug that several extra fields didn't have values when cu_diag is on (Thanks to Bart Brashers of Environ)
- Add feedback to surface diagnostic fields: T2, TH2, Q2, U10, V10 and PSFC (Thanks to Robert Fovell of UCLA)
- SKEB option: fix a problem allocating arrays before de-allocating them first (Thanks to Glenn Creighton of AFWA).
- Variable description changed for 'climate diagnostic' output variables.
- File phys/module_cu_g3.F has been corrected to allow the use of 'ishallow' with the G3 cumulus scheme, when running with WRF built with OpenMP or MPI.


- Input sounding for idealized two-fire example updated. Previous one didn't work.
- Program ndown: fix to use default hypsometric_opt = 2 option (posted 5/4/2012)
- Program real: since V3.3, the top half-level of Qv coming out of the real program (ARW only) has been zero. This is fixed in this release.
- Fix a bug associated with writing history file after restart: History output will not be written out at the time of restart. If one would like to have it, use namelist option 'write_hist_at_0h_rst = .t.' . The bug caused the behavior unpredictable. Using 'override_restart_timers' option will also work if one would like to change the output interval to a larger value. However, using the option 'override_restart_timers' may not work correctly if fdda and sst_update options are used, unless the restart time is also the time when data is available.

Experimental features:

- netCDF4 compression and chunking features. With this option and current parameter setting, the file size can be reduced up to 45%. However it is a bit clumsy to use this option in this release, unless you're running on bluefire at NCAR. It will require installation of serial version of HDF and then netCDF4. The option is activated at compile time by adding these additional macros in configure.wrf file just above the line "DMPARALLEL = 1"

NETCDFPATH = /your-netcdf4-path/lib/netcdf/4.1.3_seq
HDFPATH = /your-hdf-path/lib/hdf5/1.8.7_seq
ZLIBPATH = /your-zlib-path/lib/zlib/1.2.5

# Modify existing WRF Makefile macros

LIB_LOCAL = -L$(NETCDFPATH)/lib -lnetcdf -lnetcdff -lgpfs \
-L$(HDFPATH)/lib -lhdf5 -lhdf5_hl \
-L$(ZLIBPATH)/lib -lz

- Adding in-code radar reflectivity calculations for the following schemes:
WSM5/6, WDM5/6, Thompson, Morrison, Goddard, and Lin
The calculation is done only at output time, and can be activated via namelist option do_radar_ref = 1 in &physics. Whether you use this option or not, a new array refl_10cm will appear in wrfout file.

- pressure level diagnostics, controlled by new namelist record &diags:

p_lev_diags = 0
num_press_levels = 4
press_levels = 85000, 70000, 50000, 25000
use_tot_or_hyd_p = 2

Variables availables are: U, V, Temp, RH, geop height, wind speed and dew point temp. Output goes to auxhist23 by default. Other relevant namelists can be found in test/em_real/examples.namelist.


Previous Update

WRF V3.0.1
WRF V3.0.1.1


Home -- Model System -- User Support -- Doc / Pub -- Links -- Download -- WRF Real-time Forecast