Table of Contents

3.1   Top level of the WRF EMS distribution
3.2   Inside the bowels of the data directory
3.3   The always useful util directory

 

3.1     Top level of the WRF EMS distribution

From the top level of the WRF EMS ( $WRF ) you will see the following directories and files:

bin

data

docs

etc

gui

logs

runs

strc

util

WRF.cshrc

The bin directory contains most of the statically-linked binaries used in the WRF EMS package, including ARW and NMM core executables for both shared and distributed memory environments. Remember that all 32- and 64-bit Linux binaries are provided with the WRF EMS package so you do not need a compiler to run either core. This directory may be accessed with the $WRF_BIN environment variable.

The data directory contains the static data sets, default model configuration files, various tables, and copies of your computational domains. There are quite a few files and directories beneath the data directory with which you will need to be familiar. This directory may be accessed with the $WRF_DATA environment variable. Additional details on the contents of the data directory will be provided in Chapter 3.2.

The docs directory contains STRC, FSL, NCAR, and NCEP documentation on the WRF. Not all of the information is relevant to the STRC EMS system but users may find some value in the documentation. In addition, gif images depicting the areal coverage of the NCEP gridded data sets used to initialized model runs are provided in the docs/grids directory. This directory may be accessed with the $WRF_DOCS environment variable.

The etc directory contains NCAR/FSL Perl scripts that are used by the WRF EMS. For the most part you will not need to edit any files in this directory. This directory may be accessed with the $WRF_ETC environment variable.

The gui directory contains the Perl scripts and binaries used when running the WRF domain configuration GUI. When the WRF EMS was installed, Perl binaries were built for your system and placed in this directory. For the most part you will not need to edit any files in this directory.

The logs directory contains some of the installation and domain configuration GUI log files. The log files from your local model runs are not located here. This directory may be accessed with the $WRF_LOGS environment variable.

The runs directory contains all of your local WRF domains. This is the directory from which you will make your model runs. Once you have created a computational domain with the GUI, a directory with your selected name will be created under runs . This directory may be accessed with the $WRF_RUN environment variable.

The strc directory contains all of the Perl and Cshell scripts used to run the WRF EMS. While it is unlikely that you will ever need to edit the files contained in this directory, since they will be updated from time to time. This directory may be accessed with the $WRF_STRC environment variable.

The util directory contains the run directories for benchmark cases and various other utilities. You will need to visit this directory when benchmarking your system. There are also additional files and directories that may be of interest to the user. Additional information on the contents of the util directory is provided in Chapter 3.3. This directory may be accessed with the $WRF_UTIL environment variable.

Finally, the WRF.cshrc file contains all the environment variables used with the WRF EMS. This file will be sourced by the user when logging in.

 

3.2      Inside the bowels of the data directory

Beneath the bowels of the $WRF/data directory you will find 4 sub-directories:

geog

conf

domains

wrfsi

The geog directory contains the all static surface data sets such as topography, land use, surface albedo, etc., used during the creation of your computational domain. The files are in a binary format so you cannot look at their contents directly.

The conf directory contains three sub directories with which you should become familiar as you may need to edit some of the files contained within these directories.

The conf sub-directories are:

config

gribinfo

tables

The conf/config directory contains the default WRF model run configuration files for the WRF EMS run-time scripts contained in $WRF/strc. When you create a new computational domain with the SI GUI, the directories and files beneath $WRF/data/conf/config are copied into your new domain directory. If you want to make changes to the default configuration files then this is the location.

The conf/gribinfo directory contains files used to define all the data sets available for model initialization and boundary conditions. There is a separate <data set>_gribinfo.conf file for each available data set. It is recommended that you familiarize your self with the contents and format of these files as you may need to edit one someday. Additional information on the <data set>_gribinfo.conf files in available is Chapter 6 and appendix A.

The conf/tables directory contains tables and additional default configuration files used by the WRF EMS. There are only a few files in this directory that you may need to change.

The domains directory contains information and configuration files for each of the active WRF EMS computational domains. After installation, only files for the two benchmark cases and a general default domain will exist in domains . As you create new computational domains, your primary configuration files will be placed under the $WRF/runs directory and a backup placed in the domains directory. You should not need to edit the files in the domains directory. Should you ever want to delete a computational domain it is recommended that you use the wrf_clean utility described in Chapter 4.

 

3.3      The always useful util directory

Beneath the $WRF/util directory you will find various utilities that may be of interest to you:

bench_arw

bench_nmm

bin

gempak

grads

graphics

grib2

mpich

ncarg

The bench_arw and bench_nmm directories contain fully configured benchmark cases for the ARW and NMM cores respectively. Running the case will also serve to make sure you have everything installed correctly on your system. You are strongly encouraged to benchmark your system as one of the initial steps in setting up the WRF EMS. More information and guidance on running the benchmarks can be found in section 13.  

The bin directory contains the precompiled binaries for the various utility programs including GRADS, GEMPAK, netCDF, wgrib, and wgrib2. Enjoy!  

The gempak directory contains all the binaries, tables, and support files for creating GEMPAK formatted grid files as well as BUFKIT files. You do not need a separate NAWIPS installation on your system to create BUFKIT files as everything is included in the WRF EMS. In addition, there is a scripts sub directory that contains sample C shell scripts for generating various fields in gif format. There is a lot of good stuff in this directory with which to play around.  

The grads directory contains all the binaries, tables, and support files for using GRADS. You do not need a separate GRADS installation on your system as everything is included in the WRF EMS. In addition, there is a scripts sub directory that contains sample GRADS scripts for generating various fields in png format. There is a lot of good stuff in this directory with which to play around, just like the gempak directory.  

The graphics directory contains various NCL scripts that come with the standard ARW and NMM WRF distributions used to view the netCDF files.

The grib2 directory contains the source code for compiling grib2 for viewing grib2-formatted files. Lots of grib2 stuff.

The mpich directory contains all the MPICH (version 2.7) libraries and support files for running the WRF EMS on distributed memory systems. All the binaries have been precompiled with statically linked libraries so that there is really nothing for the user to do. The only file that may need to be edited by the user is /util/machines/machines.LINUX.

The ncarg directory contains libraries and binaries for running NCAR graphics.