Difference between revisions of "How to build WRF4"
Jump to navigation
Jump to search
(started generic) |
|||
| (7 intermediate revisions by 2 users not shown) | |||
| Line 1: | Line 1: | ||
| + | [[Category:WRF-Fire merge]] | ||
==Getting the source code== | ==Getting the source code== | ||
| − | From [https://github.com/openwfm/WRF our fork] | + | From [https://github.com/openwfm/WRF-Fire-merge our fork] |
<pre> | <pre> | ||
| − | git clone git@github.com:openwfm/WRF.git | + | git clone git@github.com:openwfm/WRF-Fire-merge.git |
</pre> | </pre> | ||
or from the [https://github.com/wrf-model/WRF official repository] | or from the [https://github.com/wrf-model/WRF official repository] | ||
| Line 8: | Line 9: | ||
git clone git@github.com:wrf-model/WRF.git | git clone git@github.com:wrf-model/WRF.git | ||
</pre> | </pre> | ||
| − | Use the <tt>master</tt> branch for the current release (default), or | + | Use the <tt>master</tt> branch for the current release (default), or branch <tt>develop</tt>, which includes changes accepted to WRF for a release in future. |
| − | < | ||
| − | |||
| − | </ | ||
| − | |||
''Coming soon:'' branch with updated fire model. | ''Coming soon:'' branch with updated fire model. | ||
| Line 21: | Line 18: | ||
cd WRF | cd WRF | ||
./configure | ./configure | ||
| − | <pre> | + | </pre> |
| − | Use <tt>./configure -d</tt> for faster build and to enable debugging. Choose serial for your platform first to verify basic functionality. Note: Currently, the fire code will not work with choices that include <tt>sm</ | + | Use <tt>./configure -d</tt> for faster build and to enable debugging. Choose <tt>serial</tt> for your platform first to verify basic functionality. Note: Currently, the fire code will not work with choices that include shared memory parallelism by OpenMP, i.e., <tt>smpar</tt> or <tt>sm+dm</tt> |
| − | + | This creates the file <tt>configure.wrf</tt>. Edit <tt>configure.wrf</tt> as needed, then | |
./compile em_fire >& compile.log | ./compile em_fire >& compile.log | ||
</pre> | </pre> | ||
| Line 37: | Line 34: | ||
./ideal.exe | ./ideal.exe | ||
./wrf.exe | ./wrf.exe | ||
| − | </ | + | </pre> |
| + | |||
| + | ==University of Utah CHPC== | ||
| + | ===Set up the environment=== | ||
| + | <pre> | ||
| + | module load intel/18 impi/18 netcdf-f | ||
| + | setenv WRFIO_NCD_LARGE_FILE_SUPPORT 1 | ||
| + | setenv NETCDF /uufs/chpc.utah.edu/sys/installdir/netcdf/i18 | ||
| + | setenv JASPERLIB /uufs/chpc.utah.edu/sys/installdir/jasper/1.900.1-atmos07102015/lib | ||
| + | setenv JASPERINC /uufs/chpc.utah.edu/sys/installdir/jasper/1.900.1-atmos07102015/include | ||
| + | </pre> | ||
| + | In <tt>./configure</tt>, select 13 - <tt>INTEL (ifort/icc)</tt> and <tt>serial</tt>. | ||
| + | ===Modify the configuration=== | ||
| + | Edit <tt>configure.wrf</tt>: Add at the end of <tt>INCLUDE_MODULES</tt> | ||
| + | <pre> | ||
| + | -I$(HDF5_INCDIR) | ||
| + | </pre> | ||
| + | and add at the end of <tt>DEP_LIB_PATH</tt> | ||
| + | <pre> | ||
| + | -L$(HDF5_LIBDIR) | ||
| + | </pre> | ||
| + | |||
| + | ===MPI parallel execution=== | ||
| + | Not tested yet. | ||
| + | ===WPS=== | ||
| + | Not tested yet. | ||
| + | ===Real runs and GeoTIFF=== | ||
| + | Not tested yet. | ||
| + | |||
| + | ==Works with== | ||
| + | * release-v4.0.2 commit [https://github.com/openwfm/WRF/commit/a9017c2de1e8430a02e5a798b756892846ac30c0 a9017c2de1e8430a02e5a798b756892846ac30c0] | ||
| + | * CHPC kingspeak11 as of January 15, 2019 | ||
Latest revision as of 19:56, 3 February 2019
Getting the source code
From our fork
git clone git@github.com:openwfm/WRF-Fire-merge.git
or from the official repository
git clone git@github.com:wrf-model/WRF.git
Use the master branch for the current release (default), or branch develop, which includes changes accepted to WRF for a release in future.
Coming soon: branch with updated fire model.
All platforms
Set up the environment as needed, then
cd WRF ./configure
Use ./configure -d for faster build and to enable debugging. Choose serial for your platform first to verify basic functionality. Note: Currently, the fire code will not work with choices that include shared memory parallelism by OpenMP, i.e., smpar or sm+dm This creates the file configure.wrf. Edit configure.wrf as needed, then ./compile em_fire >& compile.log
Check for errors and warnings
grep Err compile.log grep War compile.log
and test the code built with serial:
cd test cd em_fire ./ideal.exe ./wrf.exe
University of Utah CHPC
Set up the environment
module load intel/18 impi/18 netcdf-f setenv WRFIO_NCD_LARGE_FILE_SUPPORT 1 setenv NETCDF /uufs/chpc.utah.edu/sys/installdir/netcdf/i18 setenv JASPERLIB /uufs/chpc.utah.edu/sys/installdir/jasper/1.900.1-atmos07102015/lib setenv JASPERINC /uufs/chpc.utah.edu/sys/installdir/jasper/1.900.1-atmos07102015/include
In ./configure, select 13 - INTEL (ifort/icc) and serial.
Modify the configuration
Edit configure.wrf: Add at the end of INCLUDE_MODULES
-I$(HDF5_INCDIR)
and add at the end of DEP_LIB_PATH
-L$(HDF5_LIBDIR)
MPI parallel execution
Not tested yet.
WPS
Not tested yet.
Real runs and GeoTIFF
Not tested yet.
Works with
- release-v4.0.2 commit a9017c2de1e8430a02e5a798b756892846ac30c0
- CHPC kingspeak11 as of January 15, 2019