0% found this document useful (0 votes)

2K views

WRF Python

Python plots WRF

Uploaded by

Octavia Hopper

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

2K views

WRF Python

Python plots WRF

Uploaded by

Octavia Hopper

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 226

wrf-python Documentation

Release 1.2.0

Bill Ladwig

Sep 04, 2018

Contents

1 Documentation 3
1.1 What’s New . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.3 Table of Available Diagnostics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.4 How To Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.5 Plotting Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
1.6 API Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
1.7 Frequently Asked Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
1.8 Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
1.9 Citation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
1.10 License . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
1.11 Tutorials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195

2 Indices and tables 217

i
ii
wrf-python Documentation, Release 1.2.0

A collection of diagnostic and interpolation routines for use with output from the Weather Research and Forecasting
(WRF-ARW) Model.
This package provides over 30 diagnostic calculations, several interpolation routines, and utilities to help with plotting
via cartopy, basemap, or PyNGL. The functionality is similar to what is provided by the NCL WRF package.
When coupled with either matplotlib or PyNGL, users can create plots like this:

Contents 1
wrf-python Documentation, Release 1.2.0

2 Contents
CHAPTER 1

Documentation

1.1 What’s New

1.1.1 Releases

v1.2.0

• Release 1.2.0
• Previous versions of wrf-python promoted the strings used in xarray (e.g. name, attributes) to Unicode strings
for Python 2.7. This caused problems when porting examples for PyNGL to use wrf-python in Python 3.x. All
strings are now the native string type for the Python version being used. While this change should be transparent
to most users, any users that worked with the xarray name or attribute values on Python 2.7 may run in to string
related errors, so we’ve decided to bump the major version number.

v1.1.3

• Release 1.1.3
• Fixed/Enhanced the cloud top temperature diagnostic.
– Optical depth was not being calculated correctly when cloud ice mixing ratio was not available.
– Fixed an indexing bug that caused crashes on Windows, but should have been crashing on all plat-
forms.
– Users can now specify if they want cloud free regions to use fill values, rather than the default behavior
of using the surface temperature.
– Users can now specify the optical depth required to trigger the cloud top temperature calculation.
However, the default value of 1.0 should be sufficient for most users.
• Added ‘th’ alias for the theta product.
• Fixed a crash issue related to updraft helicity when a dictionary is used as the input.

3
wrf-python Documentation, Release 1.2.0

• Dictionary inputs now work correctly with xy_to_ll and ll_to_xy.

• The cape_2d diagnostic can now work with a single column of data, just like cape_3d.

v1.1.2

• Release 1.1.2
• Fix OpenMP directive issue with cloud top temperature.

v1.1.1

• Release 1.1.1
• Added script for building on Cheyenne with maxed out Intel settings, which also required a patch for
numpy.distutils.
• Fixed a few unicode characters hiding in a docstring that were causing problems on Cheyenne, and also building
the docs with Sphinx on Python 2.x.
• Fix issue with np.amax not working with xarray on Cheyenne, causing an error with the mdbz product.
• Fix cape_2d private variable bug when running with multiple CPUs.

v1.1.0

• Release 1.1.0
• Computational routines now support multiple cores using OpenMP. See Using OpenMP for details on how to
use this new feature.
• The CAPE routines should be noticeably faster, even in the single threaded case (thank you supreethms1809!).
• wrf.getvar() now works correctly with non-gridded NetCDF variables
• The cloud fraction diagnostic has changed:
– Users can now select their own cloud threshold levels, and can choose between a vertical coordinate
defined as height (AGL), height (MSL), or pressure.
– The default vertical coordinate type has been changed to be height (AGL). This ensures that clouds
appear over mountainous regions. If you need the old behavior, set the vert_type argument to ‘pres-
sure’.
– Fixed a bug involving the cloud threshold search algorithm, where if the surface was higher than the
threshold for a cloud level, the algorithm would use whatever was there before (uninitialized variable
bug). This caused some interesting visualization issues when plotted. Now, whenever the surface is
above a cloud level threshold, a fill value is used to indicate that data is unavailable for that location.
• The cartopy object for LambertConformal should now work correctly in the southern hemisphere.
• Fixed a bug with the PolarStereographic projection missing a geobounds argument (thank you hanschen!).
• Renamed the modules containing the ‘get_product’ routines used by wrf.getvar() to avoid naming conflicts
with the raw computational routine names. Users should be using wrf.getvar() instead of these routines,
but for those that imported the ‘get_product’ routines directly, you will need to modify your code.
• Fixed a uniqueness issue with the internal coordinate cache that was causing crashes when input data is changed
to a different file in a jupyter notebook cell.
• Added code to better support building wheels on Windows (thank you letmaik!)

4 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

• Improved support for scipy.io.netcdf objects.

• Added a new ‘zstag’ diagnostic that returns the height values for the vertically staggered grid.
• A DOI is now available for wrf-python. Please cite wrf-python if you are using it for your research. (See
Citation)
• Fixed issue with vertcross and interpline not working correctly when a projection object is used. Users will now
have to supply the lower left latitude and longitude corner point.
• Beginning with numpy 1.14, wrf-python can be built using the MSVC compiler with gfortran. WRF-Python can
now be built for Python 3.5+ on services like AppVeyor.

v1.0.5

• Release 1.0.5
• Reduced the CI test file sizes by half.

v1.0.4

• Release 1.0.4
• Fix warnings with CI tests which were caused by fill values being written as NaN to the NetCDF result file.
• Added the __eq__ operator to the WrfProj projection base class.
• Fixed array order issue when using the raw CAPE routine with 1D arrays.

v1.0.3

• Relase 1.0.3
• Fixed an issue with the cartopy Mercator subclass where the xlimits were being calculated to the same value (or
very close), causing blank plots.

v1.0.2

• Release 1.0.2
• Fixed issue with the wspd_wdir product types when sequences of files are used.

v1.0.1

• Release 1.0.1
• Fixed issue with initialization of PolarStereographic and LatLon map projection objects.
• Fixed issue where XTIME could be included in the coordinate list of a variable, but the actual XTIME variable
could be missing. NCL allows this, so wrf-python should as well.

1.1. What’s New 5

wrf-python Documentation, Release 1.2.0

v1.0.0

• Release 1.0.0.
• Fixed issue with not being able to set the thread-local coordinate cache to 0 to disable it. Also, the cache will
now correctly resize itself when the size is reduced to less than its current setting.
• Fixed an issue with the ‘0000-00-00 00:00:00’ time used in geo_em files causing crashes due to the invalid time.
The time is now set to numpy.datetime64(‘NaT’).
• Fixed issue with wrf.cape_3d not working correctly with a single column of data.

1.1.2 Beta Releases

v1.0b3

• Beta release 3.
• Improvements made for conda-forge integration testing.
• Fixed an incorrectly initialized variable issue with vinterp. This issue mainly impacts the unit tests for continu-
ous integration testing with conda-forge, since the data set used for these tests is heavily cropped.
• Back-ported the inspect.BoundArguments.apply_defaults so that Python 3.4 works. Windows users that want to
try out wrf-python with Python 3.4 can use the bladwig conda channel to get it.

v1.0b2

• Beta release 2.
• xarray 0.9 no longer includes default index dimensions in the coordinate mappings. This was causing a crash in
the routines that cause a reduction in dimension shape, mainly the interpolation routines. This has been fixed.
• Documentation updated to show the new output from xarray.

v1.0b1

• Beta release 1.
• Added more packaging boilerplate.
• Note: Currently unable to build with Python 3.5 on Windows, due to issues with distutils, numpy distutils, and
mingw compiler. Will attempt to find a workaround before the next release. Windows users should use Python
2.7 or Python 3.4 for now.

1.1.3 Alpha Releases

v1.0a3

• Alpha release 3.
• Added docstrings.
• The mapping API has changed.

6 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

– The projection attributes are no longer arrays for moving domains.

– Utility functions have been added for extracting geobounds. It is now easier to get map projection
objects from sliced variables.
– Utility functions have been added for getting cartopy, basemap, and pyngl objects.
– Users should no longer need to use xarray attributes directly
• Now uses CoordPair for cross sections so that lat/lon can be used instead of raw x,y grid coordinates.
• Renamed npvalues to to_np which is more intuitive.
• Fixed issue with generator expressions.
• Renamed some functions and arguments.

1.1.4 Known Issues

v1.0.0

• Currently unable to build on Windows with Python 3.5+ using open source mingw compiler. The mingwpy
project is working on resolving the incompatibilities between mingw and Visual Studio 2015 that was used to
build Python 3.5+. Numpy 1.13 also has improved f2py support for Python 3.5+ on Windows, so this will be
revisited when it is released.

1.2 Installation

1.2.1 Required Dependencies

• Python 2.7, 3.4, or 3.5+

• numpy (1.11 or later; 1.14 required to build on Windows)
• wrapt (1.10 or later)
• setuptools (38.0 or later)

1.2.2 Highly Recommended Packages

• xarray (0.7.0 or later)

• PyNIO (1.4.3 or later)
• netCDF4-python (1.2.0 or later)

1.2.3 Plotting Packages

• PyNGL (1.4.3 or later)

• matplotlib (1.4.3 or later)
– cartopy (0.13 or later)
– basemap (1.0.8 or later)

1.2. Installation 7
wrf-python Documentation, Release 1.2.0

1.2.4 Installing via Conda

The easiest way to install wrf-python is using Conda:

conda install -c conda-forge wrf-python

Note: If you use conda to install wrf-python on a supercomputer like Yellowstone or Cheyenne, we recommend that
you do not load any python related modules via the ‘module load’ command. The packages installed by the ‘module
load’ system will not play nicely with packages installed via conda.
Further, some systems will install python packages to a ~/.local directory, which will be found by the miniconda
python interpreter and cause various import problems. If you have a ~/.local directory, we strongly suggest renaming
it (mv ~/.local ~/.local_backup).

1.2.5 Installing on Yellowstone

On Yellowstone, wrf-python can also be installed using the module load system, if this is preferred over using conda.
Unfortunately, because wrf-python requires newer dependencies, it is not available using the ‘all-python-libs’ module,
so many of the dependencies need to be manually installed (most are for xarray).
Also, make sure you are running in the gnu/4.8.2 compiler environment or you will get import errors for a missing
libquadmath library when you go to import wrf-python.
To install:

module load gnu/4.8.2 or module swap intel gnu/4.8.2

module load python/2.7.7
module load numpy/1.11.0 wrapt/1.10.10 scipy/0.17.1 bottleneck/1.1.0 numexpr/2.6.0
˓→pyside/1.1.2 matplotlib/1.5.1 pandas/0.18.1 netcdf4python/1.2.4 xarray/0.8.2

module load wrf-python/1.0.1

1.2.6 Installing via Source Code

Installation via source code will require a Fortran and C compiler in order to run f2py. You can get them here.
The source code is available via github:
https://github.com/NCAR/wrf-python
Or PyPI:
https://pypi.python.org/pypi/wrf-python
To install, if you do not need OpenMP support, change your working directory to the wrf-python source directory and
run:

$ pip install .

Beginning with wrf-python 1.1, OpenMP is supported, but preprocessing the ompgen.F90 file is required, which also
requires running f2py to build the .pyf file. To simplify this process, you can use the build scripts in the build_scripts
directory as a guide, or just call the script directly.
Below is a sample from a build script for GNU compiler with OpenMP enabled:

8 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

cd ../fortran/build_help

gfortran -o sizes -fopenmp omp_sizes.f90

python sub_sizes.py

cd ..

gfortran -E ompgen.F90 -fopenmp -cpp -o omp.f90

f2py *.f90 -m _wrffortran -h wrffortran.pyf --overwrite-signature

cd ..

python setup.py clean --all

python setup.py config_fc --f90flags="-mtune=generic -fopenmp" build_ext --libraries=

˓→"gomp" build

pip install .

Beginning with numpy 1.14, f2py extensions can now be built using the MSVC compiler and mingw gfortran compiler.
Numpy 1.14 is required to build wrf-python for Python 3.5+.

Note: If you are building on a supercomputer and receiving linker related errors (e.g. missing symbols, undefined
references, etc), you probably need to unset the LDFLAGS environment variable. System administrators on super-
computing systems often define LDFLAGS for you so that you don’t need to worry about where libraries like NetCDF
are installed. Unfortunately, this can cause problems with the numpy.distutils build system. To fix, using the build
command from above:

$ unset LDFLAGS python setup.py config_fc --f90flags="-mtune=generic -fopenmp" build_

˓→ext --libraries="gomp" build

1.3 Table of Available Diagnostics

Variable Name Description Available Units Additional Keyword Ar-

guments
avo Absolute Vorticity 10-5 s-1
eth/theta_e Equivalent Potential Tem- K units (str) : Set to desired
perature degC units. Default is ‘K’.
degF
cape_2d 2D cape J kg-1 ; J kg-1 ; m ; m missing (float): Fill value
(mcape/mcin/lcl/lfc) for output only
cape_3d 3D cape and cin J kg-1 missing (float): Fill value
for output only
Continued on next page

1.3. Table of Available Diagnostics 9

wrf-python Documentation, Release 1.2.0

Table 1 – continued from previous page

Variable Name Description Available Units Additional Keyword Ar-
guments
ctt Cloud Top Temperature degC fill_nocloud (boolean):
K Set to True to use fill
degF values for cloud free re-
gions rather than surface
temperature. Default is
False.
missing (float): The
fill value to use when
fill_nocloud is True.
opt_thresh (float): The
optical depth required to
trigger the cloud top tem-
perature calculation. De-
fault is 1.0.
units (str) : Set to desired
units. Default is ‘degC’.
cloudfrac Cloud Fraction % vert_type (str): The
vertical coordinate type
for the cloud thresholds.
Must be ‘height_agl’,
‘height_msl’, or ‘pres’.
Default is ‘height_agl’.
low_thresh (float): The
low cloud threshold (me-
ters for ‘height_agl’ and
‘height_msl’, pascals for
‘pres’). Default is 300 m
(97000 Pa)
mid_thresh (float): The
mid cloud threshold (me-
ters for ‘height_agl’ and
‘height_msl’, pascals for
‘pres’). Default is 2000 m
(80000 Pa)
high_thresh (float): The
high cloud threshold (me-
ters for ‘height_agl’ and
‘height_msl’, pascals for
‘pres’). Default is 6000 m
(45000 Pa)
dbz Reflectivity dBZ do_variant (boolean):
Set to True to enable vari-
ant calculation. Default is
False.
do_liqskin (boolean): Set
to True to enable liquid
skin calculation. Default
is False.
Continued on next page

10 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

Table 1 – continued from previous page

Variable Name Description Available Units Additional Keyword Ar-
guments
mdbz Maximum Reflectivity dBZ do_variant (boolean):
Set to True to enable vari-
ant calculation. Default is
False.
do_liqskin (boolean): Set
to True to enable liquid
skin calculation. Default
is False.
geopt/geopotential Geopotential for the Mass m2 s-2
Grid
geopt_stag Geopotential for the Verti- m2 s-2
cally Staggered Grid
helicity Storm Relative Helicity m2 s-2 top (float): The top level
for the calculation in me-
ters. Default is 3000.0.
lat Latitude decimal degrees
lon Longitude decimal degrees
omg/omega Omega Pa s-1
p/pres Full Model Pressure Pa units (str) : Set to desired
(in specified units) hPa units. Default is ‘Pa’.
mb
torr
mmhg
atm
pressure Full Model Pressure (hPa) hPa
pvo Potential Vorticity PVU
pw Precipitable Water kg m-2
rh Relative Humidity %
rh2 2m Relative Humidity %
slp Sea Level Pressure hPa units (str) : Set to desired
hPa units. Default is ‘hPa’.
mb
torr
mmhg
atm
ter Model Terrain Height m units (str) : Set to desired
km units. Default is ‘m’.
dm
ft
mi
td2 2m Dew Point Tempera- degC units (str) : Set to desired
ture K units. Default is ‘degC’.
degF
td Dew Point Temperature degC units (str) : Set to desired
K units. Default is ‘degC’.
degF
tc Temperature in Celsius degC
Continued on next page

1.3. Table of Available Diagnostics 11

wrf-python Documentation, Release 1.2.0

Table 1 – continued from previous page

Variable Name Description Available Units Additional Keyword Ar-
guments
th/theta Potential Temperature K units (str) : Set to desired
degC units. Default is ‘K’.
degF
temp Temperature (in specified K units (str) : Set to desired
units) degC units. Default is ‘K’.
degF
tk Temperature in Kelvin K
times Times in the File or Se-
quence
xtimes XTIME Coordinate minutes since
(if applicable) start of
model run
tv Virtual Temperature K units (str) : Set to desired
degC units. Default is ‘K’.
degF
twb Wet Bulb Temperature K units (str) : Set to desired
degC units. Default is ‘K’.
degF
updraft_helicity Updraft Helicity m2 s-2 bottom (float): The bot-
tom level for the calcula-
tion in meters. Default is
2000.0.
top (float): The top level
for the calculation in me-
ters. Default is 5000.0.
ua U-component of Wind on m s-1 units (str) : Set to desired
Mass Points km h-1 units. Default is ‘m s-1’.
mi h-1
kt
ft s-1
va V-component of Wind on m s-1 units (str) : Set to desired
Mass Points km h-1 units. Default is ‘m s-1’.
mi h-1
kt
ft s-1
wa W-component of Wind on m s-1 units (str) : Set to desired
Mass Points km h-1 units. Default is ‘m s-1’.
mi h-1
kt
ft s-1
uvmet10 10 m U and V Compo- m s-1 units (str) : Set to desired
nents of Wind km h-1 units. Default is ‘m s-1’.
Rotated to Earth Coordi- mi h-1
nates kt
ft s-1
Continued on next page

12 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

Table 1 – continued from previous page

Variable Name Description Available Units Additional Keyword Ar-
guments
uvmet U and V Components of m s-1 units (str) : Set to desired
Wind km h-1 units. Default is ‘m s-1’.
Rotated to Earth Coordi- mi h-1
nates kt
ft s-1
wspd_wdir Wind Speed and Direction m s-1 units (str) : Set to desired
(wind_from_direction) km h-1 units. Default is ‘m s-1’.
in Grid Coordinates mi h-1
kt
ft s-1
wspd_wdir10 10m Wind Speed m s-1 units (str) : Set to desired
and Direction km h-1 units. Default is ‘m s-1’.
(wind_from_direction) mi h-1
in Grid Coordinates kt
ft s-1
uvmet_wspd_wdir Wind Speed and Direction m s-1 units (str) : Set to desired
(wind_from_direction) km h-1 units. Default is ‘m s-1’.
Rotated to Earth Coordi- mi h-1
nates kt
ft s-1
uvmet10_wspd_wdir 10m Wind Speed m s-1 units (str) : Set to desired
and Direction km h-1 units. Default is ‘m s-1’.
(wind_from_direction) mi h-1
Rotated to Earth Coordi- kt
nates ft s-1
z/height Model Height for Mass m msl (boolean): Set to
Grid km False to return AGL val-
dm ues. True is for MSL. De-
ft fault is True.
mi units (str) : Set to desired
units. Default is ‘m’.
zstag Model Height for Verti- m msl (boolean): Set to
cally Staggered Grid km False to return AGL val-
dm ues. True is for MSL. De-
ft fault is True.
mi units (str) : Set to desired
units. Default is ‘m’.

1.4 How To Use

1.4.1 Introduction

The API for wrf-python can be summarized as a variable computation/extraction routine, several interpolation routines,
and a few plotting helper utilities. The API is kept as simple as possible to help minimize the learning curve for new
programmers, students, and scientists. In the future, we plan to extend xarray for programmers desiring a more object
oriented API, but this remains a work in progress.
The five most commonly used routines can be summarized as:
• wrf.getvar() - Extracts WRF-ARW NetCDF variables and computes diagnostic variables that WRF does

1.4. How To Use 13

wrf-python Documentation, Release 1.2.0

not compute (e.g. storm relative helicity). This is the routine that you will use most often.
• wrf.interplevel() - Interpolates a three-dimensional field to a horizontal plane at a specified level using
simple (fast) linear interpolation (e.g. 850 hPa temperature).
• wrf.vertcross() - Interpolates a three-dimensional field to a vertical plane through a user-specified hori-
zontal line (i.e. a cross section).
• wrf.interpline() - Interpolates a two-dimensional field to a user-specified line.
• wrf.vinterp() - Interpolates a three-dimensional field to user-specified ‘surface’ levels (e.g. theta-e levels).
This is a smarter, albeit slower, version of wrf.interplevel().

1.4.2 Basic Usage

Computing Diagnostic Variables

The primary use for the wrf.getvar() function is to return diagnostic variables that require a calculation, since
WRF does not produce these variables natively. These diagnostics include CAPE, storm relative helicity, omega, sea
level pressure, etc. A table of all available diagnostics can be found here: Table of Available Diagnostics.
In the example below, sea level pressure is calculated and printed.
from __future__ import print_function

from netCDF4 import Dataset

from wrf import getvar

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Get the Sea Level Pressure

slp = getvar(ncfile, "slp")

print(slp)

Result:
<xarray.DataArray u'slp' (south_north: 1059, west_east: 1799)>
array([[ 1012.220337, 1012.298157, 1012.247864, ..., 1010.132019,
1009.932312, 1010.067078],
[ 1012.432861, 1012.444763, 1012.33667 , ..., 1010.1073 ,
1010.108459, 1010.047607],
[ 1012.395447, 1012.380859, 1012.417053, ..., 1010.22937 ,
1010.055969, 1010.026794],
...,
[ 1009.042358, 1009.069214, 1008.987793, ..., 1019.19281 ,
1019.144348, 1019.110596],
[ 1009.224854, 1009.075134, 1008.986389, ..., 1019.071899,
1019.042664, 1019.061279],
[ 1009.188965, 1009.107117, 1008.979797, ..., 1018.917786,
1018.956848, 1019.047485]], dtype=float32)
Coordinates:
XLONG (south_north, west_east) float32 -122.72 -122.693 -122.666 ...
XLAT (south_north, west_east) float32 21.1381 21.1451 21.1521 ...
Time datetime64[ns] 2016-10-07
Dimensions without coordinates: south_north, west_east
Attributes:
FieldType: 104
(continues on next page)

14 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

(continued from previous page)

MemoryOrder: XY
description: sea level pressure
units: hPa
stagger:
coordinates: XLONG XLAT
projection: LambertConformal(stand_lon=-97.5, moad_cen_lat=38.5000038147,
truelat1=38.5, truelat2=38.5, pole_lat=90.0,
pole_lon=0.0)

Extracting WRF NetCDF Variables

In addition to computing diagnostic variables (see Computing Diagnostic Variables), the wrf.getvar() function
can be used to extract regular WRF-ARW output NetCDF variables.

from future import print_function

from netCDF4 import Dataset

from wrf import getvar

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

p = getvar(ncfile, "P")

print(p)

Result:

<xarray.DataArray u'P' (bottom_top: 50, south_north: 1059, west_east: 1799)>

array([[[ 1.217539e+03, 1.225320e+03, ..., 9.876406e+02, 1.001117e+03],
[ 1.238773e+03, 1.240047e+03, ..., 1.005297e+03, 9.991719e+02],
...,
[ 9.208594e+02, 9.059141e+02, ..., 1.902922e+03, 1.904805e+03],
[ 9.172734e+02, 9.091094e+02, ..., 1.894375e+03, 1.903422e+03]],

[[ 1.219562e+03, 1.210273e+03, ..., 9.973984e+02, 9.907891e+02],

[ 1.224578e+03, 1.223508e+03, ..., 9.985547e+02, 9.921172e+02],
...,
[ 9.012734e+02, 9.052031e+02, ..., 1.897766e+03, 1.894500e+03],
[ 9.137500e+02, 9.071719e+02, ..., 1.893273e+03, 1.893664e+03]],

...,
[[ 7.233154e+00, 7.224121e+00, ..., 3.627930e+00, 3.613770e+00],
[ 7.226318e+00, 7.358154e+00, ..., 3.725098e+00, 3.634033e+00],
...,
[ 5.354248e+00, 5.406006e+00, ..., 1.282715e+01, 1.264844e+01],
[ 5.295410e+00, 5.177490e+00, ..., 1.256274e+01, 1.257642e+01]],

[[ 2.362061e+00, 2.376221e+00, ..., 1.151367e+00, 1.156982e+00],

[ 2.342529e+00, 2.403809e+00, ..., 1.198486e+00, 1.155273e+00],
...,
[ 1.732910e+00, 1.768799e+00, ..., 4.247070e+00, 4.135498e+00],
[ 1.715332e+00, 1.657227e+00, ..., 4.036377e+00, 4.047852e+00]]],
˓→dtype=float32)

Coordinates:
XLONG (south_north, west_east) float32 -122.72 -122.693 -122.666 ...
(continues on next page)

1.4. How To Use 15

wrf-python Documentation, Release 1.2.0

(continued from previous page)

XLAT (south_north, west_east) float32 21.1381 21.1451 21.1521 ...
Time datetime64[ns] 2016-10-07
Dimensions without coordinates: bottom_top, south_north, west_east
Attributes:
FieldType: 104
MemoryOrder: XYZ
description: perturbation pressure
units: Pa
stagger:
coordinates: XLONG XLAT
projection: LambertConformal(stand_lon=-97.5, moad_cen_lat=38.5000038147,
truelat1=38.5, truelat2=38.5, pole_lat=90.0,
pole_lon=0.0)

Disabling xarray and metadata

Sometimes you just want a regular numpy array and don’t care about metadata. This is often the case when you are
working with compiled extensions. Metadata can be disabled in one of two ways.
1. disable xarray completely
2. set the meta function parameter to False.
The example below illustrates both.

from future import print_function

from netCDF4 import Dataset

from wrf import getvar, disable_xarray

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Disable xarray completely

disable_xarray()
p_no_meta = getvar(ncfile, "P")
print (type(p_no_meta))
enable_xarray()

# Disable by using the meta parameter

p_no_meta = getvar(ncfile, "P", meta=False)
print (type(p_no_meta))

Result:

Extracting a Numpy Array from a DataArray

If you need to convert an xarray.DataArray to a numpy.ndarray, wrf-python provides the wrf.to_np()

function for this purpose. Although an xarray.DataArary object already contains the xarray.DataArray.
values attribute to extract the Numpy array, there is a problem when working with compiled extensions. The
behavior for xarray (and pandas) is to convert missing/fill values to NaN, which may cause crashes when working
with compiled extensions. Also, some existing code may be designed to work with numpy.ma.MaskedArray, and
numpy arrays with NaN may not work with it.

16 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

The wrf.to_np() function does the following:

1. If no missing/fill values are used, wrf.to_np() simply returns the xarray.DataArray.values at-
tribute.
2. If missing/fill values are used, then wrf.to_np() replaces the NaN values with the _FillValue found in the
xarray.DataArray.attrs attribute (required) and a numpy.ma.MaskedArray is returned.

from future import print_function

from netCDF4 import Dataset

from wrf import getvar

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Get the 3D CAPE, which contains missing values

cape_3d = getvar(ncfile, "cape_3d")

# Since there are missing values, this should return a MaskedArray

cape_3d_ndarray = to_np(cape_3d)

print(type(cape_3d_ndarray))

Result:

1.4.3 Sequences of Files

Combining Multiple Files Using the ‘cat’ Method

The ‘cat’ (concatenate) method aggregates all files in the sequence along the ‘Time’ dimension, which will be the
leftmost dimension for the output array. To include all of the times, in all of the files, in the output array, set the
timeidx parameter to wrf.ALL_TIMES (an alias for None). If a single value is specified for timeidx, then the time
index is assumed to be taken from the concatenation of all times for all files.
It is import to note that no sorting is performed in the wrf.getvar() routine, so all files in the sequence must be
sorted prior to calling this function.

from future import print_function

from netCDF4 import Dataset

from wrf import getvar, ALL_TIMES

# Creating a simple test list with three timesteps

wrflist = [Dataset("wrfout_d01_2016-10-07_00_00_00"),
Dataset("wrfout_d01_2016-10-07_01_00_00"),
Dataset("wrfout_d01_2016-10-07_02_00_00")]

# Extract the 'P' variable for all times

p_cat = getvar(wrflist, "P", timeidx=ALL_TIMES, method="cat")

print(p_cat)

Result:

1.4. How To Use 17

wrf-python Documentation, Release 1.2.0

<xarray.DataArray u'P' (Time: 3, bottom_top: 50, south_north: 1059, west_east: 1799)>

array([[[[ 1.21753906e+03, 1.22532031e+03, 1.22030469e+03, ...,
1.00760156e+03, 9.87640625e+02, 1.00111719e+03],
[ 1.23877344e+03, 1.24004688e+03, 1.22926562e+03, ...,
1.00519531e+03, 1.00529688e+03, 9.99171875e+02],
[ 1.23503906e+03, 1.23367188e+03, 1.23731250e+03, ...,
1.01739844e+03, 1.00005469e+03, 9.97093750e+02],
...,
[ 1.77978516e+00, 1.77050781e+00, 1.79003906e+00, ...,
4.22949219e+00, 4.25659180e+00, 4.13647461e+00],
[ 1.73291016e+00, 1.76879883e+00, 1.77978516e+00, ...,
4.24047852e+00, 4.24707031e+00, 4.13549805e+00],
[ 1.71533203e+00, 1.65722656e+00, 1.67480469e+00, ...,
4.06884766e+00, 4.03637695e+00, 4.04785156e+00]]]], dtype=float32)
Coordinates:
XLONG (south_north, west_east) float32 -122.72 -122.693 -122.666 ...
XLAT (south_north, west_east) float32 21.1381 21.1451 21.1521 ...
* Time (Time) datetime64[ns] 2016-10-07 2016-10-07 2016-10-07
datetime (Time) datetime64[ns] 2016-10-07T00:00:00 ...
Dimensions without coordinates: bottom_top, south_north, west_east
Attributes:
FieldType: 104
MemoryOrder: XYZ
description: perturbation pressure
units: Pa
stagger:
coordinates: XLONG XLAT
projection: LambertConformal(stand_lon=-97.5, moad_cen_lat=38.5000038147,
truelat1=38.5, truelat2=38.5, pole_lat=90.0,
pole_lon=0.0)

Combining Multiple Files Using the ‘join’ Method

The ‘join’ method combines a sequence of files by adding a new leftmost dimension for the file/sequence index. In
situations where there are multiple files with multiple times, and the last file contains less times than the previous
files, the remaining arrays will be arrays filled with missing values. There are checks in place within the wrf-python
algorithms to look for these missing arrays, but be careful when calling compiled routines outside of wrf-python.
In most cases, timeidx parameter should be set to wrf.ALL_TIMES. If a timeidx value is specified, then this time
index is used when extracting the variable from each file. In cases where there are multiple files with multiple time
steps, this is probably nonsensical, since the nth time index for each file represents a different time.
In general, join is rarely used, so the concatenate method should be used for most cases.

from future import print_function

from netCDF4 import Dataset

from wrf import getvar, ALL_TIMES

# Creating a simple test list with three timesteps

wrflist = [Dataset("wrfout_d01_2016-10-07_00_00_00"),
Dataset("wrfout_d01_2016-10-07_01_00_00"),
Dataset("wrfout_d01_2016-10-07_02_00_00")]

# Extract the 'P' variable for all times

(continues on next page)

18 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

(continued from previous page)

p_join = getvar(wrflist, "P", timeidx=ALL_TIMES, method="join")

print(p_join)

Result:

<xarray.DataArray u'P' (file: 3, bottom_top: 50, south_north: 1059, west_east: 1799)>

array([[[[ 1.217539e+03, ..., 1.001117e+03],
...,
[ 9.172734e+02, ..., 1.903422e+03]],
...,
[[ 2.362061e+00, ..., 1.156982e+00],
...,
[ 1.715332e+00, ..., 4.047852e+00]]]], dtype=float32)
Coordinates:
XLONG (south_north, west_east) float32 -122.72 -122.693 -122.666 ...
XLAT (south_north, west_east) float32 21.1381 21.1451 21.1521 ...
* file (file) int64 0 1 2
datetime (file) datetime64[ns] 2016-10-07 ...
Dimensions without coordinates: bottom_top, south_north, west_east
Attributes:
FieldType: 104
MemoryOrder: XYZ
description: perturbation pressure
units: Pa
stagger:
coordinates: XLONG XLAT
projection: LambertConformal(stand_lon=-97.5, moad_cen_lat=38.5000038147,
truelat1=38.5, truelat2=38.5, pole_lat=90.0,
pole_lon=0.0)

Note how the ‘Time’ dimension was replaced with the ‘file’ dimension, due to numpy’s automatic squeezing of the
single element ‘Time’ dimension. To maintain the ‘Time’ dimension, set the squeeze parameter to False.

from future import print_function

from netCDF4 import Dataset

from wrf import getvar, ALL_TIMES

# Creating a simple test list with three timesteps

wrflist = [Dataset("wrfout_d01_2016-10-07_00_00_00"),
Dataset("wrfout_d01_2016-10-07_01_00_00"),
Dataset("wrfout_d01_2016-10-07_02_00_00")]

# Extract the 'P' variable for all times

p_join = getvar(wrflist, "P", timeidx=ALL_TIMES, method="join", squeeze=False)

print(p_join)

Result

<xarray.DataArray u'P' (file: 3, Time: 1, bottom_top: 50, south_north: 1059, west_

˓→east: 1799)>

array([[[[[ 1.217539e+03, ..., 1.001117e+03],

...,
(continues on next page)

1.4. How To Use 19

wrf-python Documentation, Release 1.2.0

(continued from previous page)

[ 9.172734e+02, ..., 1.903422e+03]],

...,
[[ 2.362061e+00, ..., 1.156982e+00],
...,
[ 1.715332e+00, ..., 4.047852e+00]]]]], dtype=float32)
Coordinates:
XLONG (south_north, west_east) float32 -122.72 -122.693 -122.666 ...
XLAT (south_north, west_east) float32 21.1381 21.1451 21.1521 ...
* file (file) int64 0 1 2
datetime (file, Time) datetime64[ns] 2016-10-07 2016-10-07 2016-10-07
Dimensions without coordinates: Time, bottom_top, south_north, west_east
Attributes:
FieldType: 104
MemoryOrder: XYZ
description: perturbation pressure
units: Pa
stagger:
coordinates: XLONG XLAT
projection: LambertConformal(stand_lon=-97.5, moad_cen_lat=38.5000038147,
truelat1=38.5, truelat2=38.5, pole_lat=90.0,
pole_lon=0.0)

Dictionaries of WRF File Sequences

Dictionaries can also be used as input to the wrf.getvar() functions. This can be useful when working with
ensembles. However, all WRF files in the dictionary must have the same dimensions. The result is an array where the
leftmost dimension is the keys from the dictionary. Nested dictionaries are allowed.
The method argument is used to describe how each sequence in the dictionary will be combined.

from future import print_function

from netCDF4 import Dataset

from wrf import getvar, ALL_TIMES

wrf_dict = {"ens1" : [Dataset("ens1/wrfout_d01_2016-10-07_00_00_00"),

Dataset("ens1/wrfout_d01_2016-10-07_01_00_00"),
Dataset("ens1/wrfout_d01_2016-10-07_02_00_00")],
"ens2" : [Dataset("ens2/wrfout_d01_2016-10-07_00_00_00"),
Dataset("ens2/wrfout_d01_2016-10-07_01_00_00"),
Dataset("ens2/wrfout_d01_2016-10-07_02_00_00")]
}

p = getvar(wrf_dict, "P", timeidx=ALL_TIMES)

print(p)

Result:

<xarray.DataArray 'P' (key_0: 2, Time: 2, bottom_top: 50, south_north: 1059, west_

˓→east: 1799)>

array([[[[[ 1.217539e+03, ..., 1.001117e+03],

...,
[ 9.172734e+02, ..., 1.903422e+03]],
(continues on next page)

20 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

(continued from previous page)

...,
[[ 2.362061e+00, ..., 1.156982e+00],
...,
[ 1.715332e+00, ..., 4.047852e+00]]]]], dtype=float32)
Coordinates:
XLONG (south_north, west_east) float32 -122.72 -122.693 -122.666 ...
XLAT (south_north, west_east) float32 21.1381 21.1451 21.1521 ...
* Time (Time) datetime64[ns] 2016-10-07 ...
datetime (Time) datetime64[ns] 2016-10-07 ...
* key_0 (key_0) <U6 u'label1' u'label2'
Dimensions without coordinates: bottom_top, south_north, west_east
Attributes:
FieldType: 104
MemoryOrder: XYZ
description: perturbation pressure
units: Pa
stagger:
coordinates: XLONG XLAT
projection: LambertConformal(stand_lon=-97.5, moad_cen_lat=38.5000038147,
truelat1=38.5, truelat2=38.5, pole_lat=90.0,
pole_lon=0.0)

1.4.4 Interpolation Routines

Interpolating to a Horizontal Level

The wrf.interplevel() function is used to interpolate a 3D field to a specific horizontal level, usually pressure
or height.

from future import print_function

from netCDF4 import Dataset

from wrf import getvar, interplevel

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Extract the Geopotential Height and Pressure (hPa) fields

z = getvar(ncfile, "z")
p = getvar(ncfile, "pressure")

# Compute the 500 MB Geopotential Height

ht_500mb = interplevel(z, p, 500.)

print(ht_500mb)

Result:

<xarray.DataArray u'height_500_hPa' (south_north: 1059, west_east: 1799)>

array([[ 5882.16992188, 5881.87939453, 5881.81005859, ...,
5890.14501953, 5890.23583984, 5890.33349609],
[ 5882.71777344, 5882.17529297, 5882.1171875 , ...,
5890.37695312, 5890.38525391, 5890.27978516],
[ 5883.32177734, 5882.47119141, 5882.34130859, ...,
(continues on next page)

1.4. How To Use 21

wrf-python Documentation, Release 1.2.0

(continued from previous page)

5890.48339844, 5890.42871094, 5890.17724609],
...,
[ 5581.45800781, 5580.46826172, 5579.32617188, ...,
5788.93554688, 5788.70507812, 5788.64453125],
[ 5580.32714844, 5579.51611328, 5578.34863281, ...,
5788.15869141, 5787.87304688, 5787.65527344],
[ 5579.64404297, 5578.30957031, 5576.98632812, ...,
5787.19384766, 5787.10888672, 5787.06933594]], dtype=float32)
Coordinates:
XLONG (south_north, west_east) float32 -122.72 -122.693 -122.666 ...
XLAT (south_north, west_east) float32 21.1381 21.1451 21.1521 ...
Time datetime64[ns] 2016-10-07
Dimensions without coordinates: south_north, west_east
Attributes:
FieldType: 104
units: m
stagger:
coordinates: XLONG XLAT
projection: LambertConformal(stand_lon=-97.5, moad_cen_lat=38.5000038147,
truelat1=38.5, truelat2=38.5, pole_lat=90.0,
pole_lon=0.0)
level: 500 hPa
missing_value: 9.96920996839e+36
_FillValue: 9.96920996839e+36

Vertical Cross Sections

The wrf.vertcross() function is used to create vertical cross sections. To define a cross section, a start point
and an end point needs to be specified. Alternatively, a pivot point and an angle may be used. The start point, end
point, and pivot point are specified using a wrf.CoordPair object, and coordinates can either be in grid (x,y)
coordinates or (latitude,longitude) coordinates. When using (latitude,longitude) coordinates, a NetCDF file object or
a wrf.WrfProj object must be provided.
The vertical levels can also be specified using the levels parameter. If not specified, then approximately 100 levels will
be chosen in 1% increments.

Example Using Start Point and End Point

from future import print_function, division

from netCDF4 import Dataset

from wrf import getvar, vertcross, CoordPair

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Get the geopotential height (m) and pressure (hPa).

z = getvar(ncfile, "z")
p = getvar(ncfile, "pressure")

# Define a start point and end point in grid coordinates

start_point = CoordPair(x=0, y=(z.shape[-2]-1)//2)
end_point = CoordPair(x=-1, y=(z.shape[-2]-1)//2)

(continues on next page)

22 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

(continued from previous page)

# Calculate the vertical cross section. By setting latlon to True, this
# also calculates the latitude and longitude coordinates along the cross
# section line and adds them to the 'xy_loc' metadata to help with plotting.
p_vert = vertcross(p, z, start_point=start_point, end_point=end_point, latlon=True)

print(p_vert)

Result:

<xarray.DataArray u'pressure_cross' (vertical: 100, idx: 1798)>

array([[ nan, nan, nan, ..., nan,
nan, nan],
[ 989.66168213, 989.66802979, 989.66351318, ..., 988.05737305,
987.99151611, 987.96917725],
[ 959.49450684, 959.50109863, 959.50030518, ..., 958.96948242,
958.92980957, 958.89294434],
...,
[ 24.28092003, 24.27359581, 24.27034378, ..., 24.24800491,
24.2486496 , 24.24947357],
[ 23.2868309 , 23.27933884, 23.27607918, ..., 23.25231361,
23.2530098 , 23.25384521],
[ nan, nan, nan, ..., nan,
nan, nan]], dtype=float32)
Coordinates:
Time datetime64[ns] 2016-10-07
xy_loc (idx) object CoordPair(x=0.0, y=529.0, lat=34.5279502869, lon=-127.
˓→398925781) ...

* vertical (vertical) float32 0.0 261.828 523.656 785.484 1047.31 1309.14 ...
Dimensions without coordinates: idx
Attributes:
FieldType: 104
description: pressure
units: hPa
stagger:
coordinates: XLONG XLAT
projection: LambertConformal(stand_lon=-97.5, moad_cen_lat=38.5000038147,
truelat1=38.5, truelat2=38.5, pole_lat=90.0,
pole_lon=0.0)
orientation: (0.0, 529.0) to (1797.0, 529.0)
missing_value: 9.96920996839e+36
_FillValue: 9.96920996839e+36

Example Using Pivot Point and Angle

from future import print_function, division

from netCDF4 import Dataset

from wrf import getvar, vertcross, CoordPair

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Get the geopotential height (m) and pressure (hPa).

z = getvar(ncfile, "z")
p = getvar(ncfile, "pressure")
(continues on next page)

1.4. How To Use 23

wrf-python Documentation, Release 1.2.0

(continued from previous page)

# Define a pivot point and angle in grid coordinates, with the

# pivot point being the center of the grid.
pivot_point = CoordPair(x=(z.shape[-1]-1)//2, y=(z.shape[-2]-1)//2)
angle = 90.0

# Calculate the vertical cross section. By setting latlon to True, this

# also calculates the latitude and longitude coordinates along the line
# and adds them to the metadata to help with plotting labels.
p_vert = vertcross(p, z, pivot_point=pivot_point, angle=angle, latlon=True)

print (p_vert)

Result:

<xarray.DataArray u'pressure_cross' (vertical: 100, idx: 1798)>

* vertical (vertical) float32 0.0 261.828 523.656 785.484 1047.31 1309.14 ...
Dimensions without coordinates: idx
Attributes:
FieldType: 104
description: pressure
units: hPa
stagger:
coordinates: XLONG XLAT
projection: LambertConformal(stand_lon=-97.5, moad_cen_lat=38.5000038147,
truelat1=38.5, truelat2=38.5, pole_lat=90.0,
pole_lon=0.0)
orientation: (0.0, 529.0) to (1797.0, 529.0) ; center=CoordPair(x=899.0, y=529.0)
˓→; angle=90.0

missing_value: 9.96920996839e+36
_FillValue: 9.96920996839e+36

Example Using Lat/Lon Coordinates

from future import print_function, division

from netCDF4 import Dataset

(continues on next page)

24 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

(continued from previous page)

from wrf import getvar, vertcross, CoordPair

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Get the geopotential height (m) and pressure (hPa).

z = getvar(ncfile, "z")
p = getvar(ncfile, "pressure")
lats = getvar(ncfile, "lat")
lons = getvar(ncfile, "lon")

# Making the same horizontal line, but with lats/lons

start_lat = lats[(lats.shape[-2]-1)//2, 0]
end_lat = lats[(lats.shape[-2]-1)//2, -1]
start_lon = lons[(lats.shape[-2]-1)//2, 0]
end_lon = lons[(lats.shape[-2]-1)//2, -1]

# Cross section line using start_point and end_point.

start_point = CoordPair(lat=start_lat, lon=start_lon)
end_point = CoordPair(lat=end_lat, lon=end_lon)

# When using lat/lon coordinates, you must supply a WRF netcdf file object,
# or a projection object with the lower left latitude and lower left
# longitude points.
p_vert = vertcross(p, z, wrfin=ncfile, start_point=start_point, end_point=end_point,
˓→latlon=True)

print(p_vert)

Result:
<xarray.DataArray u'pressure_cross' (vertical: 100, idx: 1798)>
array([[ nan, nan, nan, ..., nan,
nan, nan],
[ 989.66168213, 989.66802979, 989.66351318, ..., 988.05737305,
987.99151611, 987.96917725],
[ 959.49450684, 959.50109863, 959.50030518, ..., 958.96948242,
958.92980957, 958.89294434],
...,
[ 24.28092003, 24.27359581, 24.27034378, ..., 24.24800491,
24.2486496 , 24.24947357],
[ 23.2868309 , 23.27933884, 23.27607918, ..., 23.25231361,
23.2530098 , 23.25384521],
[ nan, nan, nan, ..., nan,
nan, nan]], dtype=float32)
Coordinates:
Time datetime64[ns] 2016-10-07
xy_loc (idx) object CoordPair(x=0.0, y=529.0, lat=34.5279502869, lon=-127.
˓→398925781) ...

1.4. How To Use 25

wrf-python Documentation, Release 1.2.0

(continued from previous page)

pole_lon=0.0)
orientation: (0.0, 529.0) to (1797.0, 529.0)
missing_value: 9.96920996839e+36
_FillValue: 9.96920996839e+36

Example Using Specified Vertical Levels

from future import print_function, division

from netCDF4 import Dataset

from wrf import getvar, vertcross, CoordPair

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Get the geopotential height (m) and pressure (hPa).

z = getvar(ncfile, "z")
p = getvar(ncfile, "pressure")
lats = getvar(ncfile, "lat")
lons = getvar(ncfile, "lon")

# Making the same horizontal line, but with lats/lons

start_lat = lats[(lats.shape[-2]-1)//2, 0]
end_lat = lats[(lats.shape[-2]-1)//2, -1]
start_lon = lons[(lats.shape[-2]-1)//2, 0]
end_lon = lons[(lats.shape[-2]-1)//2, -1]

# Pressure using start_point and end_point. These were obtained using

start_point = CoordPair(lat=start_lat, lon=start_lon)
end_point = CoordPair(lat=end_lat, lon=end_lon)

# Specify vertical levels

levels = [1000., 2000., 3000.]

# Calculate the cross section

p_vert = vertcross(p, z, wrfin=ncfile, levels=levels, start_point=start_point, end_
˓→point=end_point, latlon=True)

print(p_vert)

Result:

<xarray.DataArray u'pressure_cross' (vertical: 3, idx: 1798)>

array([[ 906.375 , 906.38043213, 906.39367676, ..., 907.6661377 ,
907.63006592, 907.59191895],
[ 804.24737549, 804.26885986, 804.28076172, ..., 806.98632812,
806.95556641, 806.92608643],
[ 713.24578857, 713.2722168 , 713.27886963, ..., 716.09594727,
716.06610107, 716.03503418]], dtype=float32)
Coordinates:
Time datetime64[ns] 2016-10-07
xy_loc (idx) object CoordPair(x=0.0, y=529.0, lat=34.5279502869, lon=-127.
˓→398925781) ...

* vertical (vertical) float32 1000.0 2000.0 3000.0

Dimensions without coordinates: idx
(continues on next page)

26 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

(continued from previous page)

Attributes:
FieldType: 104
description: pressure
units: hPa
stagger:
coordinates: XLONG XLAT
projection: LambertConformal(stand_lon=-97.5, moad_cen_lat=38.5000038147,
truelat1=38.5, truelat2=38.5, pole_lat=90.0,
pole_lon=0.0)
orientation: (0.0, 529.0) to (1797.0, 529.0)
missing_value: 9.96920996839e+36
_FillValue: 9.96920996839e+36

Interpolating Two-Dimensional Fields to a Line

Two-dimensional fields can be interpolated along a line, in a manner similar to the vertical cross section (see Vertical
Cross Sections), using the wrf.interpline() function. To define the line to interpolate along, a start point and
an end point needs to be specified. Alternatively, a pivot point and an angle may be used. The start point, end
point, and pivot point are specified using a wrf.CoordPair object, and coordinates can either be in grid (x,y)
coordinates or (latitude,longitude) coordinates. When using (latitude,longitude) coordinates, a NetCDF file object or
a wrf.WrfProj object must also be provided.

Example Using Start Point and End Point

from future import print_function, division

from netCDF4 import Dataset

from wrf import getvar, interpline, CoordPair

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Get the 2m temperature

t2 = getvar(ncfile, "T2")

# Create a south-north line in the center of the domain using

# start point and end point
start_point = CoordPair(x=(t2.shape[-1]-1)//2, y=0)
end_point = CoordPair(x=(t2.shape[-1]-1)//2, y=-1)

# Calculate the vertical cross section. By setting latlon to True, this

# also calculates the latitude and longitude coordinates along the line
# and adds them to the metadata to help with plotting labels.
t2_line = interpline(t2, start_point=start_point, end_point=end_point, latlon=True)

print(t2_line, "\n")

Result:
<xarray.DataArray u'T2_line' (line_idx: 1058)>
array([ 302.07214355, 302.08505249, 302.08688354, ..., 279.18557739,
279.1998291 , 279.23132324], dtype=float32)
Coordinates:
Time datetime64[ns] 2016-10-07
(continues on next page)

1.4. How To Use 27

wrf-python Documentation, Release 1.2.0

(continued from previous page)

xy_loc (line_idx) object CoordPair(x=899.0, y=0.0, lat=24.3645858765, lon=-97.
˓→5) ...

Dimensions without coordinates: line_idx

Attributes:
FieldType: 104
description: TEMP at 2 M
units: K
stagger:
coordinates: XLONG XLAT
projection: LambertConformal(stand_lon=-97.5, moad_cen_lat=38.5000038147,
truelat1=38.5, truelat2=38.5, pole_lat=90.0,
pole_lon=0.0)
orientation: (899.0, 0.0) to (899.0, 1057.0)

Example Using Pivot Point and Angle

from future import print_function, division

from netCDF4 import Dataset

from wrf import getvar, interpline, CoordPair

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Get the 2m temperature

t2 = getvar(ncfile, "T2")

# Create a south-north line using pivot point and angle

pivot_point = CoordPair((t2.shape[-1]-1)//2, (t2.shape[-2]-1)//2)
angle = 0.0

# Calculate the vertical cross section. By setting latlon to True, this

# also calculates the latitude and longitude coordinates along the line
# and adds them to the metadata to help with plotting labels.
t2_line = interpline(t2, pivot_point=pivot_point, angle=angle, latlon=True)

print(t2_line, "\n")

Result:

<xarray.DataArray u'T2_line' (line_idx: 1058)>

array([ 302.07214355, 302.08505249, 302.08688354, ..., 279.18557739,
279.1998291 , 279.23132324], dtype=float32)
Coordinates:
Time datetime64[ns] 2016-10-07
xy_loc (line_idx) object CoordPair(x=899.0, y=0.0, lat=24.3645858765, lon=-97.
˓→5) ...

Dimensions without coordinates: line_idx

Attributes:
FieldType: 104
description: TEMP at 2 M
units: K
stagger:
coordinates: XLONG XLAT
projection: LambertConformal(stand_lon=-97.5, moad_cen_lat=38.5000038147,
(continues on next page)

28 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

(continued from previous page)

truelat1=38.5, truelat2=38.5, pole_lat=90.0,
pole_lon=0.0)
orientation: (899.0, 0.0) to (899.0, 1057.0) ; center=CoordPair(x=899, y=529) ;
˓→angle=0.0

Example Using Lat/Lon Coordinates

from future import print_function, division

from netCDF4 import Dataset

from wrf import getvar, interpline, CoordPair

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

t2 = getvar(ncfile, "T2")
lats = getvar(ncfile, "lat")
lons = getvar(ncfile, "lon")

# Select the latitude,longitude points for a vertical line through

# the center of the domain.
start_lat = lats[0, (lats.shape[-1]-1)//2]
end_lat = lats[-1, (lats.shape[-1]-1)//2]
start_lon = lons[0, (lons.shape[-1]-1)//2]
end_lon = lons[-1, (lons.shape[-1]-1)//2]

# Create the CoordPairs

start_point = CoordPair(lat=start_lat, lon=start_lon)
end_point = CoordPair(lat=end_lat, lon=end_lon)

# Calculate the interpolated line. To use latitude and longitude points,

# you must supply a WRF NetCDF file object, or a projection object along
# with the lower left latitude and lower left longitude points.
# Also, by setting latlon to True, this routine will
# also calculate the latitude and longitude coordinates along the line
# and adds them to the metadata to help with plotting labels.
t2_line = interpline(t2, wrfin=ncfile, start_point=start_point, end_point=end_point,
˓→latlon=True)

print (t2_line)

Result:

<xarray.DataArray u'T2_line' (line_idx: 1058)>

Dimensions without coordinates: line_idx

Attributes:
FieldType: 104
description: TEMP at 2 M
units: K
(continues on next page)

1.4. How To Use 29

wrf-python Documentation, Release 1.2.0

(continued from previous page)

stagger:
coordinates: XLONG XLAT
projection: LambertConformal(stand_lon=-97.5, moad_cen_lat=38.5000038147,
truelat1=38.5, truelat2=38.5, pole_lat=90.0,
pole_lon=0.0)
orientation: (899.0, 0.0) to (899.0, 1057.0)

Interpolating a 3D Field to a Surface Type

The wrf.vinterp() is used to interpolate a field to a type of surface. The available surfaces are pressure, geopo-
tential height, theta, and theta-e. The surface levels to interpolate also need to be specified.

from future import print_function

from netCDF4 import Dataset

from wrf import getvar, vinterp

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

tk = getvar(ncfile, "tk")
# Interpolate tk to theta-e levels
interp_levels = [200, 300, 500, 1000]

interp_field = vinterp(ncfile,
field=tk,
vert_coord="eth",
interp_levels=interp_levels,
extrapolate=True,
field_type="tk",
log_p=True)

print(interp_field)

Result:

<xarray.DataArray u'temp' (interp_level: 4, south_north: 1059, west_east: 1799)>

array([[[ 296.12872314, 296.1166687 , 296.08905029, ..., 301.71026611,
301.67956543, 301.67791748],
[ 296.11352539, 295.95581055, 295.91555786, ..., 301.63052368,
301.62905884, 301.65887451],
[ 296.07556152, 295.91577148, 295.88214111, ..., 301.61499023,
301.60287476, 301.63961792],
...,
[ 219.11134338, 219.08581543, 219.08602905, ..., 218.29879761,
218.30923462, 218.3787384 ],
[ 219.09260559, 219.07765198, 219.08340454, ..., 218.2855072 ,
218.30444336, 218.37931824],
[ 219.07936096, 219.08181763, 219.10089111, ..., 218.31173706,
218.34288025, 218.3687439 ]]], dtype=float32)
Coordinates:
XLONG (south_north, west_east) float32 -122.72 -122.693 -122.666 ...
XLAT (south_north, west_east) float32 21.1381 21.1451 21.1521 ...
Time datetime64[ns] 2016-10-07
* interp_level (interp_level) int64 200 300 500 1000
Dimensions without coordinates: south_north, west_east
(continues on next page)

30 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

(continued from previous page)

Attributes:
FieldType: 104
MemoryOrder: XYZ
description: temperature
units: K
stagger:
coordinates: XLONG XLAT
projection: LambertConformal(stand_lon=-97.5, moad_cen_lat=38.5000038147,
truelat1=38.5, truelat2=38.5, pole_lat=90.0,
pole_lon=0.0)
vert_interp_type: eth

1.4.5 Lat/Lon <-> XY Routines

wrf-python includes a set of routines for converting back and forth between latitude,longitude space and x,y space. The
methods are wrf.xy_to_ll(), wrf.xy_to_ll_proj(), wrf.ll_to_xy(), wrf.ll_to_xy_proj().
The latitude, longitude, x, and y parameters to these methods can contain sequences if multiple points are desired to
be converted.

Example With Single Coordinates

from future import print_function

from netCDF4 import Dataset

from wrf import getvar, interpline, CoordPair, xy_to_ll, ll_to_xy

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

lat_lon = xy_to_ll(ncfile, 400, 200)

print(lat_lon)

x_y = ll_to_xy(ncfile, lat_lon[0], lat_lon[1])

print (x_y)

Result:

<xarray.DataArray u'latlon' (lat_lon: 2)>

array([ 28.55816408, -112.67827617])
Coordinates:
* lat_lon (lat_lon) <U3 u'lat' u'lon'
xy_coord object CoordPair(x=400, y=200)
Dimensions without coordinates: idx

<xarray.DataArray u'xy' (x_y: 2)>

array([400, 200])
Coordinates:
latlon_coord object CoordPair(lat=28.5581640822, lon=-112.678276173)
* x_y (x_y) <U1 u'x' u'y'
Dimensions without coordinates: idx

1.4. How To Use 31

wrf-python Documentation, Release 1.2.0

Example With Multiple Coordinates

from future import print_function

from netCDF4 import Dataset

from wrf import getvar, interpline, CoordPair, xy_to_ll, ll_to_xy

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

lat_lon = xy_to_ll(ncfile, [400,105], [200,205])

print(lat_lon)

x_y = ll_to_xy(ncfile, lat_lon[0,:], lat_lon[1,:])

print (x_y)

Result:

<xarray.DataArray u'latlon' (lat_lon: 2, idx: 2)>

array([[ 28.55816408, 27.03835783],
[-112.67827617, -121.36392174]])
Coordinates:
* lat_lon (lat_lon) <U3 u'lat' u'lon'
xy_coord (idx) object CoordPair(x=400, y=200) CoordPair(x=105, y=205)
Dimensions without coordinates: idx

<xarray.DataArray u'xy' (x_y: 2, idx: 2)>

array([[400, 105],
[200, 205]])
Coordinates:
latlon_coord (idx) object CoordPair(lat=28.5581640822, lon=-112.678276173) ...
* x_y (x_y) <U1 u'x' u'y'
Dimensions without coordinates: idx

1.4.6 Mapping Helper Routines

wrf-python includes several routines to assist with plotting, primarily for obtaining the mapping object used for car-
topy, basemap, and PyNGL. For all three plotting systems, the mapping object can be determined directly from a
variable when using xarray, or can be obtained from the WRF output file(s) if xarray is turned off.
Also included are utilities for extracting the geographic boundaries directly from xarray variables. This can be useful
in situations where you only want to work with subsets (slices) of a large domain, but don’t want to define the map
projection over the subset region.

Cartopy Example Using a Variable

In this example, we’re going to extract the cartopy mapping object from a diagnostic variable (slp), the lat,lon coor-
dinates, and the geographic boundaries. Next, we’re going to take a subset of the diagnostic variable and extract the
geographic boundaries. Some of the variables will be printed for demonstration.

from future import print_function

(continues on next page)

32 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

(continued from previous page)

from netCDF4 import Dataset
from wrf import getvar, get_cartopy, latlon_coords, geo_bounds

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Use SLP for the example variable

slp = getvar(ncfile, "slp")

# Get the cartopy mapping object

cart_proj = get_cartopy(slp)

print (cart_proj)

# Get the latitude and longitude coordinate. This is usually needed for plotting.
lats, lons = latlon_coords(slp)

# Get the geobounds for the SLP variable

bounds = geo_bounds(slp)

print (bounds)

# Get the geographic boundaries for a subset of the domain

slp_subset = slp[150:250, 150:250]
slp_subset_bounds = geo_bounds(slp_subset)

print (slp_subset_bounds)

Result:

<cartopy.crs.LambertConformal object at 0x115374290>

GeoBounds(CoordPair(lat=25.9246292114, lon=-119.675048828), CoordPair(lat=29.
˓→0761833191, lon=-117.46484375))

GeoBounds(CoordPair(lat=25.9246292114, lon=-119.675048828), CoordPair(lat=29.

˓→0761833191, lon=-117.46484375))

Cartopy Example Using WRF Output Files

In this example, the cartopy mapping object and geographic boundaries will be extracted directly from the netcdf
variable.

from future import print_function

from netCDF4 import Dataset

from wrf import get_cartopy, geo_bounds

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Get the cartopy mapping object from the netcdf file

cart_proj = get_cartopy(wrfin=ncfile)

print (cart_proj)

# Get the geobounds from the netcdf file (by default, uses XLAT, XLONG)
# You can supply a variable name to get the staggered boundaries
bounds = geo_bounds(wrfin=ncfile)
(continues on next page)

1.4. How To Use 33

wrf-python Documentation, Release 1.2.0

(continued from previous page)

print (bounds)

Result:

<cartopy.crs.LambertConformal object at 0x11d3be650>

GeoBounds(CoordPair(lat=21.1381225586, lon=-122.719528198), CoordPair(lat=47.
˓→8436355591, lon=-60.9013671875))

Basemap Example Using a Variable

In this example, we’re going to extract the basemap mapping object from a diagnostic variable (slp), the lat,lon coor-
dinates, and the geographic boundaries. Next, we’re going to take a subset of the diagnostic variable and extract the
geographic boundaries. Some of the variables will be printed for demonstration.

from future import print_function

from netCDF4 import Dataset

from wrf import getvar, get_basemap, latlon_coords, geo_bounds

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

slp = getvar(ncfile, "slp")

# Get the basemap mapping object

bm = get_basemap(slp)

print (bm)

# Get the latitude and longitude coordinate. This is usually needed for plotting.
lats, lons = latlon_coords(slp)

# Get the geobounds for the SLP variable

bounds = geo_bounds(slp)

print(bounds)

# Get the geographic boundaries for a subset of the domain

slp_subset = slp[150:250, 150:250]
slp_subset_bounds = geo_bounds(slp_subset)

print (slp_subset_bounds)

Result:

<mpl_toolkits.basemap.Basemap object at 0x114d65650>

GeoBounds(CoordPair(lat=21.1381225586, lon=-122.719528198), CoordPair(lat=47.
˓→8436355591, lon=-60.9013671875))

GeoBounds(CoordPair(lat=25.9246292114, lon=-119.675048828), CoordPair(lat=29.

˓→0761833191, lon=-117.46484375)

34 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

Basemap Example Using WRF Output Files

In this example, the basemap mapping object and geographic boundaries will be extracted directly from the netcdf
variable.

from future import print_function

from netCDF4 import Dataset

from wrf import get_basemap, geo_bounds

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Get the basemap object from the netcdf file

bm = get_basemap(wrfin=ncfile)

print (bm)

# Get the geographic boundaries from the netcdf file

bounds = geo_bounds(wrfin=ncfile)

print (bounds)

Result:

<mpl_toolkits.basemap.Basemap object at 0x125bb4750>

GeoBounds(CoordPair(lat=21.1381225586, lon=-122.719528198), CoordPair(lat=47.
˓→8436355591, lon=-60.9013671875))

PyNGL Example Using a Variable

from future import print_function

from netCDF4 import Dataset

from wrf import getvar, get_pyngl, latlon_coords, geo_bounds

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Use SLP as the example variable

slp = getvar(ncfile, "slp")

# Get the pyngl resources from the variable

pyngl_resources = get_pyngl(slp)

print (pyngl_resources)

# Get the latitude and longitude coordinate. This is needed for plotting.
lats, lons = latlon_coords(slp)

# Get the geobounds from the SLP variable

bounds = geo_bounds(slp)

print(bounds)
(continues on next page)

1.4. How To Use 35

wrf-python Documentation, Release 1.2.0

(continued from previous page)

# Get the geographic boundaries for a subset of the domain

slp_subset = slp[150:250, 150:250]
slp_subset_bounds = geo_bounds(slp_subset)

print (slp_subset_bounds)

Result:

<Ngl.Resources instance at 0x114cabbd8>

GeoBounds(CoordPair(lat=21.1381225586, lon=-122.719528198), CoordPair(lat=47.
˓→8436355591, lon=-60.9013671875))

GeoBounds(CoordPair(lat=25.9246292114, lon=-119.675048828), CoordPair(lat=29.

˓→0761833191, lon=-117.46484375))

PyNGL Example Using WRF Output Files

In this example, the basemap mapping object and geographic boundaries will be extracted directly from the netcdf
variable.

from future import print_function

from netCDF4 import Dataset

from wrf import get_pyngl, geo_bounds

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Get the pyngl resources from the netcdf file

pyngl_resources = get_pyngl(wrfin=ncfile)

print (pyngl_resources)

# Get the geographic boundaries from the netcdf file

bounds = geo_bounds(wrfin=ncfile)

print (bounds)

Result:

<Ngl.Resources instance at 0x115391f80>

GeoBounds(CoordPair(lat=21.1381225586, lon=-122.719528198), CoordPair(lat=47.
˓→8436355591, lon=-60.9013671875))

Moving Nests

When a domain nest is moving, the domain boundaries become a function of time when combining the files using
the ‘cat’ method. When using ‘join’, the domain boundaries become a function of both file and time. As a result, the
methods that depend on geographic boundaries (wrf.geo_bounds(), wrf.get_basemap(), etc) will return
arrays of objects rather than a single object when multiple times and/or files are detected in the underlying coordinate
variables.
An exception is wrf.get_cartopy(), which contains no geographic boundary information in the mapping ob-
ject. Instead, the wrf.cartopy_xlim() and wrf.cartopy_ylim() methods can be used to get the array of
matplotlib axes boundaries (returned in the axes projection coordinates).

36 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

Geographic Boundaries with Moving Nest Example

In this example, the geographic boundaries are extracted from a sequence of files that use a moving nest. The result
will be an array of wrf.GeoBounds objects.

from future import print_function

from glob import glob

from netCDF4 import Dataset as nc

from wrf import getvar, ALL_TIMES, geo_bounds

# Get all the domain 02 files

wrf_filenames = glob("wrf_files/wrf_vortex_multi/wrfout_d02_*")
ncfiles = [nc(x) for x in wrf_filenames]

# SLP is the example variable and includes all times

slp = getvar(ncfiles, "slp", timeidx=ALL_TIMES)

# Get the geographic boundaries

bounds = geo_bounds(slp)
print (bounds)

Result:

[ GeoBounds(CoordPair(lat=21.3020038605, lon=-90.5740585327), CoordPair(lat=29.

˓→0274410248, lon=-82.0291671753))

GeoBounds(CoordPair(lat=21.3020038605, lon=-90.3042221069), CoordPair(lat=29.

˓→0274410248, lon=-81.7593231201))

GeoBounds(CoordPair(lat=21.3020038605, lon=-90.8438949585), CoordPair(lat=29.

˓→0274410248, lon=-82.2990036011))

GeoBounds(CoordPair(lat=21.3020038605, lon=-91.1137390137), CoordPair(lat=29.

˓→0274410248, lon=-82.5688400269))

GeoBounds(CoordPair(lat=21.8039493561, lon=-91.6534042358), CoordPair(lat=29.

˓→4982528687, lon=-83.1085205078))

GeoBounds(CoordPair(lat=22.0542640686, lon=-92.193107605), CoordPair(lat=29.

˓→7328338623, lon=-83.6481933594))

GeoBounds(CoordPair(lat=22.5535621643, lon=-92.7327728271), CoordPair(lat=30.

˓→2003688812, lon=-84.1878738403))

GeoBounds(CoordPair(lat=22.8025398254, lon=-93.0026092529), CoordPair(lat=30.

˓→4333114624, lon=-84.4577102661))

GeoBounds(CoordPair(lat=23.0510597229, lon=-93.2724456787), CoordPair(lat=30.

˓→665681839, lon=-84.7275543213))]

Cartopy Mapping with Moving Nest Example

In this example, a cartopy mapping object is extracted from a variable that uses a moving nest. Since cartopy objects
do not include geographic boundary information, only a single cartopy object is returned. However, if the axes xlimits
and ylimits are desired, the wrf.cartopy_xlim() and wrf.cartopy_ylim() functions can be used to obtain
the array of moving boundaries in the axes projected coordinate space.

from future import print_function

from glob import glob

from netCDF4 import Dataset as nc
(continues on next page)

1.4. How To Use 37

wrf-python Documentation, Release 1.2.0

(continued from previous page)

from wrf import getvar, ALL_TIMES, get_cartopy, cartopy_xlim, cartopy_ylim

# Get all of the domain 02 WRF output files

wrf_filenames = glob("wrf_files/wrf_vortex_multi/wrfout_d02_*")
ncfiles = [nc(x) for x in wrf_filenames]

# Use SLP as the example variable and include all times

slp = getvar(ncfiles, "slp", timeidx=ALL_TIMES)

# Get the cartopy mapping object

cart_proj = get_cartopy(slp)
print (cart_proj)
print ("\n")

# Get the array of axes x-limits

xlims = cartopy_xlim(slp)
print (xlims)
print ("\n")

# Get the array of axes y-limits

ylims = cartopy_ylim(slp)
print (ylims)

Result:
<wrf.projection.MercatorWithLatTS object at 0x13893c9b0>

[[-174999.8505754546, 774999.5806103835]
[-145000.11853874932, 805000.1608638937]
[-204999.58261215844, 744999.8485736783]
[-235000.16286567, 715000.1165369744]
[-294998.77872227144, 654999.804246759]
[-355001.6356629085, 595000.34017335]
[-415000.25151950994, 535000.0278831345]
[-444999.98355621524, 505000.29584642925]
[-474999.7155929191, 474999.7155929177]]

[[2424828.507236154, 3374828.14098255]
[2424828.507236154, 3374828.14098255]
[2424828.507236154, 3374828.14098255]
[2424828.507236154, 3374828.14098255]
[2484829.1182174017, 3434828.972518358]
[2514829.1041871575, 3464828.196283651]
[2574829.0041584675, 3524828.8880928173]
[2604829.1786526926, 3554829.5610342724]
[2634828.9016262344, 3584828.016406863]]

Basemap Mapping with Moving Nest Example

In this example, basemap objects are extracted from a variable that uses a moving nest. An array of basemap objects
is returned because the basemap object includes geographic boundary information.
from __future__ import print_function

(continues on next page)

38 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

(continued from previous page)

from glob import glob
from netCDF4 import Dataset as nc

from wrf import getvar, ALL_TIMES, get_basemap

# Get all of the domain 02 WRF output files

wrf_filenames = glob("wrf_files/wrf_vortex_multi/wrfout_d02_*")
ncfiles = [nc(x) for x in wrf_filenames]

# Use SLP as the reference variable and include all times

slp = getvar(ncfiles, "slp", timeidx=ALL_TIMES)

# Get the array of basemap objects

bm = get_basemap(slp)
print (bm)
print ("\n")

Result:

[<mpl_toolkits.basemap.Basemap object at 0x1327bc510>

<mpl_toolkits.basemap.Basemap object at 0x115a9a790>
<mpl_toolkits.basemap.Basemap object at 0x115a9a750>
<mpl_toolkits.basemap.Basemap object at 0x115a9a7d0>
<mpl_toolkits.basemap.Basemap object at 0x115a9a850>
<mpl_toolkits.basemap.Basemap object at 0x115a9a8d0>
<mpl_toolkits.basemap.Basemap object at 0x115a9a950>
<mpl_toolkits.basemap.Basemap object at 0x115a9a9d0>
<mpl_toolkits.basemap.Basemap object at 0x115a9aa50>]

PyNGL Mapping with Moving Nest Example

In this example, pyngl resource objects are extracted from a variable that uses a moving nest. An array of pyngl
resource objects is returned because the pyngl object includes geographic boundary information.

from future import print_function

from glob import glob

from netCDF4 import Dataset as nc

from wrf import getvar, ALL_TIMES, get_pyngl

# Get the domain 02 WRF output files

wrf_filenames = glob("wrf_files/wrf_vortex_multi/wrfout_d02_*")
ncfiles = [nc(x) for x in wrf_filenames]

# Use SLP as the example variable and include all times

slp = getvar(ncfiles, "slp", timeidx=ALL_TIMES)

# Get the array of pyngl resource objects

bm = get_pyngl(slp)
print (bm)
print ("\n")

Result:

1.4. How To Use 39

wrf-python Documentation, Release 1.2.0

[<Ngl.Resources instance at 0x140cd30e0>

<Ngl.Resources instance at 0x11d3187a0>
<Ngl.Resources instance at 0x11d3185a8>
<Ngl.Resources instance at 0x11d3188c0>
<Ngl.Resources instance at 0x11d318878>
<Ngl.Resources instance at 0x11d3183f8>
<Ngl.Resources instance at 0x11d318950>
<Ngl.Resources instance at 0x11d318a70>
<Ngl.Resources instance at 0x11d318710>]

1.4.7 Using OpenMP

Beginning in version 1.1, the Fortran computational routines in wrf-python make use of OpenMP directives. OpenMP
enables the calculations to use multiple CPU cores, which can improve performance. In order to use OpenMP features,
wrf-python has to be compiled with OpenMP enabled (most pre-built binary installations will have this enabled).
The Fortran computational routines have all been built using runtime scheduling, instead of compile time scheduling,
so that the user can choose the scheduler type within their Python application. By default, the scheduling type is set to
wrf.OMP_SCHED_STATIC using only 1 CPU core, so wrf-python will behave similarly to the non-OpenMP built
versions. For the most part, the difference between the scheduling types is minimal, with the exception being the
wrf.OMP_SCHED_DYNAMIC scheduler that is much slower due to the additional overhead associated with it. For
new users, using the default scheduler should be sufficient.

Verifying that OpenMP is Enabled

To take advantage of the performance improvements offered by OpenMP, wrf-python needs to have been compiled
with OpenMP features enabled. The example below shows how you can determine if OpenMP is enabled in your build
of wrf-python.

from future import print_function

from wrf import omp_enabled

print(omp_enabled())

Result:

True

Determining the Number of Available Processors

The example below shows how you can get the maximum number of processors that are available on your system.

from future import print_function

from wrf import omp_get_num_procs

print(omp_get_num_procs())

Result:

40 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

Specifying the Number of Threads

To enable multicore support via OpenMP, specifying the maximum number of OpenMP threads (i.e. CPU cores) is
the only step that you need to take.
In the example below, wrf.omp_set_num_threads() is used to set the maximum number of threads to use, and
wrf.omp_get_max_threads() is used to retrieve (and print) the maximum number of threads used.

Note: Although there is an OpenMP routine named wrf.omp_get_num_threads(), this routine will always
return 1 when called from the sequential part of the program. Use wrf.omp_get_max_threads() to return the
value set by wrf.omp_set_num_threads().

from future import print_function

from wrf import omp_set_num_threads, omp_get_max_threads

omp_set_num_threads(4)

print (omp_get_max_threads())

Result:

Setting a Different Scheduler Type

When an OpenMP directive is encountered in the Fortran code, a scheduler is used to determine how the work is
divided among the threads. All of the Fortran routines are compiled to use a ‘runtime’ scheduler, which indicates that
the scheduler type (from the four listed below) is to be chosen at runtime (i.e. inside a Python script)
By default, the scheduler chosen is the wrf.OMP_SCHED_STATIC scheduler, which should be sufficient for most
users. However, OpenMP and wrf-python include the following options for the scheduler type:
• wrf.OMP_SCHED_STATIC
• wrf.OMP_SCHED_DYNAMIC
• wrf.OMP_SCHED_GUIDED
• wrf.OMP_SCHED_AUTO
Refer to the OpenMP Specification. for more information about these scheduler types. In local testing, wrf.
OMP_SCHED_GUIDED produced the best results, but differences between wrf.OMP_SCHED_STATIC, wrf.
OMP_SCHED_GUIDED, and wrf.OMP_SCHED_AUTO were minor. However, wrf.OMP_SCHED_DYNAMIC pro-
duced noticeably slower results due to the overhead of using a dynamic scheduler.
When setting a scheduler type, the wrf.omp_set_schedule() takes two arguments. The first is the sched-
uler type (one from the list above), and the second optional argument is a modifier, which is usually referred as the
chunk size. If the modifier/chunk_size is set to 0, then the OpenMP default implementation is used. For wrf.
OMP_SCHED_AUTO, the modifier is ignored.
If you are new to OpenMP and all this sounds confusing, don’t worry about setting a scheduler type. The default static
scheduler will be good enough.
In the example below, the scheduler type is set to wrf.OMP_SCHED_GUIDED and uses the default chunk size of 0.
The scheduler type is then read back using wrf.omp_get_schedule() and printed.

1.4. How To Use 41

wrf-python Documentation, Release 1.2.0

from future import print_function

from wrf import omp_set_schedule, omp_get_schedule, OMP_SCHED_GUIDED

omp_set_schedule(OMP_SCHED_GUIDED, 0)

sched, modifier = omp_get_schedule()

print(sched, modifier)

Result:

3 1

Notice that the printed scheduler type (sched variable) is set to a value of 3, which is the actual integer constant value
for the wrf.OMP_SCHED_GUIDED scheduler type. The modifier is returned as a value of 1, which is different than
the 0 that was supplied to the wrf.omp_set_schedule() routine. This is because the 0 tells OpenMP to use its
own default value for the scheduler, which is 1 for this type of scheduler.

1.4.8 Performance Tips

Memory Issues with wrf.ALL_TIMES

The use of wrf.ALL_TIMES for the timeidx parameter to wrf.getvar() is convenient for computing diagnostic
variables across multiple files/times, but there is something that users should be aware of. When wrf.ALL_TIMES
is set as the timeidx argument, all arrays used in the computation are extracted for all times before the computation is
started. This can cause serious memory issues on smaller hardware systems like laptops.
In this example, the user wants to use a data set that is 289 x 39 x 300 x 300 and compute z for the entire data set. The
user is using a laptop with 8 GB of memory.

from netCDF4 import Dataset

from wrf import getvar, ALL_TIMES

file_list = [Dataset("/path/to/file1"), Dataset("/path/to/file2"),...]

z = getvar(file_list, "z", ALL_TIMES)

Five hours later, the computation finished. What happened?

In wrf-python, all of the computational routines use 8-byte REAL variables so that both the 4-byte and 8-byte version
of WRF output can be used. The calculation for z extracts three variables (P, PHB, and HGT) and returns a fourth
array (RESULT). The RESULT will get cut in half to 4-byte REALs after the computation, but needs an 8-byte REAL
when the result is computed.
Let’s look at the approximate amount memory needed:
P: 289 x 39 x 300 x 300 x 8 = 8,115,120,000 bytes (~8 GB!)
PHB: 289 x 39 x 300 x 300 x 8 = 8,115,120,000 bytes (~8 GB!)
HGT: 289 x 300 x 300 x 8 = 208,080,000 (~208 MB)
RESULT: 289 x 39 x 300 x 300 x 8 = 8,115,120,000 bytes (~8 GB!)
Yikes! So, in order to do this calculation using wrf.ALL_TIMES as the timeidx, over 24.2 GB are needed for this
one calculation. When the laptop runs out of memory, it begins using the hard drive for swap memory, which runs
hundreds of times slower than real memory.

42 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

To fix this situation, it is better to allocate the output array yourself and run the calculation for each time step in a loop
(“loop-and-fill”). The required memory requirements change to:
(Note: only need to store the result in a 4-byte REAL)
FINAL_RESULT: 289 x 39 x 300 x 300 x 4 = 4,057560,000 bytes (~4 GB)
(Note: the numbers below are for each loop iteration)
P: 39 x 300 x 300 x 8 = 28,080,000 bytes (~28 MB)
PHB: 39 x 300 x 300 x 8 = 28,080,000 bytes (~28 MB)
HGT: 300 x 300 x 8 = 720,000 bytes (720 KB)
RESULT: 39 x 300 x 300 x 8 = 28,080,000 bytes (~28 MB)
Since the memory for the computation is deleted after each loop iteration, the total memory usage drops to approxi-
mately 4.1 GB.
The moral of the story is that you need to make sure that your system has enough memory to extract everything it
needs up front if you want to use wrf.ALL_TIMES, otherwise it is better to “loop-and-fill” yourself.
Here is an example of the “loop-and-fill” technique:

from future import print_function, division

import numpy as np
from netCDF4 import Dataset
from wrf import getvar, ALL_TIMES

filename_list = ["/path/to/file1", "/path/to/file2",...]

# Result shape (hard coded for this example)

result_shape = (289, 39, 300, 300)

# Only need 4-byte floats

z_final = np.empty(result_shape, np.float32)

# Modify this number if using more than 1 time per file

times_per_file = 1

for timeidx in range(result_shape[0]):

# Compute the file index and the time index inside the file
fileidx = timeidx // times_per_file
file_timeidx = timeidx % times_per_file

f = Dataset(filename_list[fileidx])
z = getvar(f, "z", file_timeidx)

z_final[timeidx,:] = z[:]
f.close()

The cache Argument for wrf.getvar()

If you have read through the documentation, you may have noticed that the wrf.getvar() routine contains a cache
argument. What is this for?
Internally, if metadata is turned on, a variable is extracted from the NetCDF file and its metadata is copied to form
the result’s metadata. Often this variable is one of the computation’s function arguments, so rather than spend time

1.4. How To Use 43

wrf-python Documentation, Release 1.2.0

extracting the variable again for the computation, it is placed in a cache (dictionary) and passed on to the computational
function.
What isn’t widely known is that this cache argument can also be supplied by end users wishing to speed up their
application. This can be useful in situations where numerous calculations are being performed on the same data set.
For many algorithms, the time needed to extract the arrays from the NetCDF file is on par with the time needed to
perform the calculation. If you are computing numerous diagnostics, extracting the variables up front allows you to
only pay this extraction penalty once, rather than inside of each call to wrf.getvar().
The cache is nothing more than a dictionary where each key is the variable name (e.g. “P”) and the value is
the xarray.DataArray or numpy.ndarray variable. Creating the cache dictionary is easy, since the wrf.
extract_vars() routine returns a dictionary for a sequence of variables.

Note: The timeidx parameter supplied to extract_vars() must be the same timeidx parameter that you plan to
use for wrf.getvar(). Otherwise, it will crash with dimension mismatch errors.

Some common variables that you can use to create an effective cache are: P, PB, PH, PHB, T, QVAPOR, HGT, PSFC,
U, V, W.
Below is an example showing the same computation done with and without the cache. The execution time is printed.
The hardware used is a 2.8 GHz Intel Core i7, which contains 4 CPU cores with 2 hyper threads (8 total threads). This
will be interpreted as 8 CPUs for OpenMP.
from __future__ import print_function

import time
from netCDF4 import Dataset
from wrf import getvar, ALL_TIMES, extract_vars

# The first two files contain four times, the last file contains only one.
wrf_filenames = ["/path/to/wrfout_d02_2005-08-28_00:00:00",
"/path/to/wrfout_d02_2005-08-28_12:00:00",
"/path/to/wrfout_d02_2005-08-29_00:00:00"]

wrfin = [Dataset(x) for x in wrf_filenames]

start = time.time()
my_cache = extract_vars(wrfin, ALL_TIMES, ("P", "PSFC", "PB", "PH", "PHB",
"T", "QVAPOR", "HGT", "U", "V",
"W"))
end = time.time()
print ("Time taken to build cache: ", (end-start), "s")

vars = ("avo", "eth", "cape_2d", "cape_3d", "ctt", "dbz", "mdbz",

"geopt", "helicity", "lat", "lon", "omg", "p", "pressure",
"pvo", "pw", "rh2", "rh", "slp", "ter", "td2", "td", "tc",
"theta", "tk", "tv", "twb", "updraft_helicity", "ua", "va",
"wa", "uvmet10", "uvmet", "z", "cfrac", "zstag", "geopt_stag")

# No cache
start = time.time()
for var in vars:
v = getvar(wrfin, var, ALL_TIMES)
end = time.time()
no_cache_time = (end-start)

print ("Time taken without variable cache: ", no_cache_time, "s")

(continues on next page)

44 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

(continued from previous page)

# With a cache
start = time.time()
for var in vars:
v = getvar(wrfin, var, ALL_TIMES, cache=my_cache)
end = time.time()
cache_time = (end-start)

print ("Time taken with variable cache: ", cache_time, "s")

improvement = ((no_cache_time-cache_time)/no_cache_time) * 100

print ("The cache decreased computation time by: ", improvement, "%")

Result:
Time taken to build cache: 0.28154706955 s
Time taken without variable cache: 11.0905270576 s
Time taken with variable cache: 8.25931215286 s
The cache decreased computation time by: 25.5282268378 %

By removing the repeated extraction of common variables in the wrf.getvar() routine, for the single threaded
case, the computation time has been reduced by 25.5% in this particular example.
Things get more interesting when OpenMP is turned on and set to use the maximum number of processors (in this
case 8 threads are used).
from __future__ import print_function

import time
from netCDF4 import Dataset
from wrf import (getvar, ALL_TIMES, extract_vars,
omp_set_num_threads, omp_get_num_procs)

wrfin = [Dataset(x) for x in wrf_filenames]

omp_set_num_threads(omp_get_num_procs())

vars = ("avo", "eth", "cape_2d", "cape_3d", "ctt", "dbz", "mdbz",

# No cache
start = time.time()
(continues on next page)

1.4. How To Use 45

wrf-python Documentation, Release 1.2.0

(continued from previous page)

for var in vars:
v = getvar(wrfin, var, ALL_TIMES)
end = time.time()
no_cache_time = (end-start)

print ("Time taken without variable cache: ", no_cache_time, "s")

# With a cache
start = time.time()
for var in vars:
v = getvar(wrfin, var, ALL_TIMES, cache=my_cache)
end = time.time()
cache_time = (end-start)

print ("Time taken with variable cache: ", cache_time, "s")

improvement = ((no_cache_time-cache_time)/no_cache_time) * 100

print ("The cache decreased computation time by: ", improvement, "%")

Result:

Time taken to build cache: 0.2700548172 s

Time taken without variable cache: 6.02652812004 s
Time taken with variable cache: 3.27777099609 s
The cache decreased computation time by: 45.6109565772 %

In this example, 4 CPU cores (8 total threads) are used. When the cache is used, the computation time drops by 45%,
so almost half the time was spent simply extracting variables from the NetCDF file. When compared to the 11.09
s needed to compute the single threaded case with no variable cache, the computation time drops by roughly 70%
(compared to 45% with 8 threads but no cache).
In summary, if you are computing a lot of diagnostic variables, consider using the cache argument to improve perfor-
mance, particularly if you want to maximize your multithreaded performance with OpenMP.

1.5 Plotting Examples

The examples below show how wrf-python can be used to make plots with matplotlib (with basemap and cartopy) and
PyNGL. None of these examples make use of xarray’s builtin plotting functions, since additional work is most likely
needed to extend xarray in order to work correctly. This is planned for a future release.

1.5.1 Matplotlib With Cartopy

Cartopy is becoming the main tool for base mapping with matplotlib, but you should be aware of a few shortcomings
when working with WRF data.
• The builtin transformations of coordinates when calling the contouring functions do not work correctly with the
rotated pole projection. The transform_points method needs to be called manually on the latitude and longitude
arrays.
• The rotated pole projection requires the x and y limits to be set manually using set_xlim and set_ylim.
• You can’t place latitude and longitude labels on the axes when using any projection other than Mercator or
LatLon.

46 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

Plotting a Two-dimensional Field

from netCDF4 import Dataset

import matplotlib.pyplot as plt
from matplotlib.cm import get_cmap
import cartopy.crs as crs
from cartopy.feature import NaturalEarthFeature

from wrf import to_np, getvar, smooth2d, get_cartopy, cartopy_xlim, cartopy_ylim,

˓→latlon_coords

# Open the NetCDF file

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Get the sea level pressure

slp = getvar(ncfile, "slp")

# Smooth the sea level pressure since it tends to be noisy near the mountains
smooth_slp = smooth2d(slp, 3)

# Get the latitude and longitude points

lats, lons = latlon_coords(slp)

# Get the cartopy mapping object

cart_proj = get_cartopy(slp)

# Create a figure
fig = plt.figure(figsize=(12,9))
# Set the GeoAxes to the projection used by WRF
ax = plt.axes(projection=cart_proj)

# Download and add the states and coastlines

states = NaturalEarthFeature(category='cultural', scale='50m', facecolor='none',
(continues on next page)

1.5. Plotting Examples 47

wrf-python Documentation, Release 1.2.0

(continued from previous page)

name='admin_1_states_provinces_shp')
ax.add_feature(states, linewidth=.5)
ax.coastlines('50m', linewidth=0.8)

# Make the contour outlines and filled contours for the smoothed sea level pressure.
plt.contour(to_np(lons), to_np(lats), to_np(smooth_slp), 10, colors="black",
transform=crs.PlateCarree())
plt.contourf(to_np(lons), to_np(lats), to_np(smooth_slp), 10, transform=crs.
˓→PlateCarree(),

cmap=get_cmap("jet"))

# Add a color bar

plt.colorbar(ax=ax, shrink=.62)

# Set the map limits. Not really necessary, but used for demonstration.
ax.set_xlim(cartopy_xlim(smooth_slp))
ax.set_ylim(cartopy_ylim(smooth_slp))

# Add the gridlines

ax.gridlines(color="black", linestyle="dotted")

plt.title("Sea Level Pressure (hPa)")

plt.show()

48 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

Horizontal Interpolation to a Pressure Level

from netCDF4 import Dataset

import numpy as np
import matplotlib.pyplot as plt
from matplotlib.cm import get_cmap
import cartopy.crs as crs
from cartopy.feature import NaturalEarthFeature

from wrf import getvar, interplevel, to_np, latlon_coords, get_cartopy, cartopy_xlim,

˓→cartopy_ylim

# Open the NetCDF file

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Extract the pressure, geopotential height, and wind variables

p = getvar(ncfile, "pressure")
z = getvar(ncfile, "z", units="dm")
ua = getvar(ncfile, "ua", units="kt")
va = getvar(ncfile, "va", units="kt")
wspd = getvar(ncfile, "wspd_wdir", units="kts")[0,:]

# Interpolate geopotential height, u, and v winds to 500 hPa

ht_500 = interplevel(z, p, 500)
u_500 = interplevel(ua, p, 500)
v_500 = interplevel(va, p, 500)
(continues on next page)

1.5. Plotting Examples 49

wrf-python Documentation, Release 1.2.0

(continued from previous page)

wspd_500 = interplevel(wspd, p, 500)

# Get the lat/lon coordinates

lats, lons = latlon_coords(ht_500)

# Get the map projection information

cart_proj = get_cartopy(ht_500)

# Create the figure

fig = plt.figure(figsize=(12,9))
ax = plt.axes(projection=cart_proj)

# Download and add the states and coastlines

states = NaturalEarthFeature(category='cultural', scale='50m', facecolor='none',
name='admin_1_states_provinces_shp')
ax.add_feature(states, linewidth=0.5)
ax.coastlines('50m', linewidth=0.8)

# Add the 500 hPa geopotential height contours

levels = np.arange(520., 580., 6.)
contours = plt.contour(to_np(lons), to_np(lats), to_np(ht_500), levels=levels, colors=
˓→"black",

transform=crs.PlateCarree())
plt.clabel(contours, inline=1, fontsize=10, fmt="%i")

# Add the wind speed contours

levels = [25, 30, 35, 40, 50, 60, 70, 80, 90, 100, 110, 120]
wspd_contours = plt.contourf(to_np(lons), to_np(lats), to_np(wspd_500), levels=levels,
cmap=get_cmap("rainbow"),
transform=crs.PlateCarree())
plt.colorbar(wspd_contours, ax=ax, orientation="horizontal", pad=.05)

# Add the 500 hPa wind barbs, only plotting every 125th data point.
plt.barbs(to_np(lons[::125,::125]), to_np(lats[::125,::125]), to_np(u_500[::125,
˓→::125]),

to_np(v_500[::125, ::125]), transform=crs.PlateCarree(), length=6)

# Set the map bounds

ax.set_xlim(cartopy_xlim(ht_500))
ax.set_ylim(cartopy_ylim(ht_500))

ax.gridlines()

plt.title("500 MB Height (dm), Wind Speed (kt), Barbs (kt)")

plt.show()

Panel Plots From Front Page

This lengthy example shows how to make the panel plots on the first page of the documentation. For a simpler example
of how to make a cross section plot, see Vertical Cross Section.

50 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

import numpy as np
import matplotlib.pyplot as plt
from matplotlib.cm import get_cmap
import cartopy.crs as crs
import cartopy.feature as cfeature
from netCDF4 import Dataset

from wrf import (getvar, to_np, vertcross, smooth2d, CoordPair, GeoBounds, get_
˓→cartopy,

latlon_coords, cartopy_xlim, cartopy_ylim)

# Open the NetCDF file

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Get the WRF variables

slp = getvar(ncfile, "slp")
smooth_slp = smooth2d(slp, 3)
ctt = getvar(ncfile, "ctt")
z = getvar(ncfile, "z")
dbz = getvar(ncfile, "dbz")
Z = 10**(dbz/10.)
wspd = getvar(ncfile, "wspd_wdir", units="kt")[0,:]
(continues on next page)

1.5. Plotting Examples 51

wrf-python Documentation, Release 1.2.0

(continued from previous page)

# Set the start point and end point for the cross section
start_point = CoordPair(lat=26.76, lon=-80.0)
end_point = CoordPair(lat=26.76, lon=-77.8)

# Compute the vertical cross-section interpolation. Also, include the lat/lon

# points along the cross-section in the metadata by setting latlon to True.
z_cross = vertcross(Z, z, wrfin=ncfile, start_point=start_point, end_point=end_point,
latlon=True, meta=True)
wspd_cross = vertcross(wspd, z, wrfin=ncfile, start_point=start_point, end_point=end_
˓→point,

latlon=True, meta=True)
dbz_cross = 10.0 * np.log10(z_cross)

# Get the lat/lon points

lats, lons = latlon_coords(slp)

# Get the cartopy projection object

cart_proj = get_cartopy(slp)

# Create a figure that will have 3 subplots

fig = plt.figure(figsize=(10,7))
ax_ctt = fig.add_subplot(1,2,1,projection=cart_proj)
ax_wspd = fig.add_subplot(2,2,2)
ax_dbz = fig.add_subplot(2,2,4)

# Download and create the states, land, and oceans using cartopy features
states = cfeature.NaturalEarthFeature(category='cultural', scale='50m', facecolor=
˓→'none',

name='admin_1_states_provinces_shp')
land = cfeature.NaturalEarthFeature(category='physical', name='land', scale='50m',
facecolor=cfeature.COLORS['land'])
ocean = cfeature.NaturalEarthFeature(category='physical', name='ocean', scale='50m',
facecolor=cfeature.COLORS['water'])

# Make the pressure contours

contour_levels = [960, 965, 970, 975, 980, 990]
c1 = ax_ctt.contour(lons, lats, to_np(smooth_slp), levels=contour_levels, colors=
˓→"white",

transform=crs.PlateCarree(), zorder=3, linewidths=1.0)

# Create the filled cloud top temperature contours

contour_levels = [-80.0, -70.0, -60, -50, -40, -30, -20, -10, 0, 10]
ctt_contours = ax_ctt.contourf(to_np(lons), to_np(lats), to_np(ctt), contour_levels,
cmap=get_cmap("Greys"), transform=crs.PlateCarree(),
˓→zorder=2)

ax_ctt.plot([start_point.lon, end_point.lon], [start_point.lat, end_point.lat],

color="yellow", marker="o", transform=crs.PlateCarree(), zorder=3)

# Create the color bar for cloud top temperature

cb_ctt = fig.colorbar(ctt_contours, ax=ax_ctt, shrink=.60)
cb_ctt.ax.tick_params(labelsize=5)

# Draw the oceans, land, and states

ax_ctt.add_feature(ocean)
ax_ctt.add_feature(land)
(continues on next page)

52 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

(continued from previous page)

ax_ctt.add_feature(states, linewidth=.5, edgecolor="black")

# Crop the domain to the region around the hurricane

hur_bounds = GeoBounds(CoordPair(lat=np.amin(to_np(lats)), lon=-85.0),
CoordPair(lat=30.0, lon=-72.0))
ax_ctt.set_xlim(cartopy_xlim(ctt, geobounds=hur_bounds))
ax_ctt.set_ylim(cartopy_ylim(ctt, geobounds=hur_bounds))
ax_ctt.gridlines(color="white", linestyle="dotted")

# Make the contour plot for wind speed

wspd_contours = ax_wspd.contourf(to_np(wspd_cross), cmap=get_cmap("jet"))
# Add the color bar
cb_wspd = fig.colorbar(wspd_contours, ax=ax_wspd)
cb_wspd.ax.tick_params(labelsize=5)

# Make the contour plot for dbz

levels = [5 + 5*n for n in range(15)]
dbz_contours = ax_dbz.contourf(to_np(dbz_cross), levels=levels, cmap=get_cmap("jet"))
cb_dbz = fig.colorbar(dbz_contours, ax=ax_dbz)
cb_dbz.ax.tick_params(labelsize=5)

# Set the x-ticks to use latitude and longitude labels

coord_pairs = to_np(dbz_cross.coords["xy_loc"])
x_ticks = np.arange(coord_pairs.shape[0])
x_labels = [pair.latlon_str() for pair in to_np(coord_pairs)]
ax_wspd.set_xticks(x_ticks[::20])
ax_wspd.set_xticklabels([], rotation=45)
ax_dbz.set_xticks(x_ticks[::20])
ax_dbz.set_xticklabels(x_labels[::20], rotation=45, fontsize=4)

# Set the y-ticks to be height

vert_vals = to_np(dbz_cross.coords["vertical"])
v_ticks = np.arange(vert_vals.shape[0])
ax_wspd.set_yticks(v_ticks[::20])
ax_wspd.set_yticklabels(vert_vals[::20], fontsize=4)
ax_dbz.set_yticks(v_ticks[::20])
ax_dbz.set_yticklabels(vert_vals[::20], fontsize=4)

# Set the x-axis and y-axis labels

ax_dbz.set_xlabel("Latitude, Longitude", fontsize=5)
ax_wspd.set_ylabel("Height (m)", fontsize=5)
ax_dbz.set_ylabel("Height (m)", fontsize=5)

# Add a title
ax_ctt.set_title("Cloud Top Temperature (degC)", {"fontsize" : 7})
ax_wspd.set_title("Cross-Section of Wind Speed (kt)", {"fontsize" : 7})
ax_dbz.set_title("Cross-Section of Reflectivity (dBZ)", {"fontsize" : 7})

plt.show()

1.5.2 Matplotlib with Basemap

Although basemap is in maintenance mode only and becoming deprecated, it is still widely used by many program-
mers. Cartopy is becoming the preferred package for mapping, however it suffers from growing pains in some areas
(can’t use latitude/longitude labels for many map projections). If you run in to these issues, basemap is likely to

1.5. Plotting Examples 53

wrf-python Documentation, Release 1.2.0

accomplish what you need.

Plotting a Two-Dimensional Field

from netCDF4 import Dataset

import matplotlib.pyplot as plt
from matplotlib.cm import get_cmap
from mpl_toolkits.basemap import Basemap

from wrf import to_np, getvar, smooth2d, get_basemap, latlon_coords

# Open the NetCDF file

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Get the sea level pressure

slp = getvar(ncfile, "slp")

# Smooth the sea level pressure since it tends to be noisy near the mountains
smooth_slp = smooth2d(slp, 3)

# Get the latitude and longitude points

lats, lons = latlon_coords(slp)

# Get the basemap object

bm = get_basemap(slp)

# Create a figure
fig = plt.figure(figsize=(12,9))

# Add geographic outlines

bm.drawcoastlines(linewidth=0.25)
bm.drawstates(linewidth=0.25)
(continues on next page)

54 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

(continued from previous page)

bm.drawcountries(linewidth=0.25)

# Convert the lats and lons to x and y. Make sure you convert the lats and lons to
# numpy arrays via to_np, or basemap crashes with an undefined RuntimeError.
x, y = bm(to_np(lons), to_np(lats))

# Draw the contours and filled contours

bm.contour(x, y, to_np(smooth_slp), 10, colors="black")
bm.contourf(x, y, to_np(smooth_slp), 10, cmap=get_cmap("jet"))

# Add a color bar

plt.colorbar(shrink=.62)

plt.title("Sea Level Pressure (hPa)")

plt.show()

Horizontal Interpolation to a Pressure Level

from netCDF4 import Dataset

import numpy as np
import matplotlib.pyplot as plt
from matplotlib.cm import get_cmap
(continues on next page)

1.5. Plotting Examples 55

wrf-python Documentation, Release 1.2.0

(continued from previous page)

from wrf import getvar, interplevel, to_np, get_basemap, latlon_coords

# Open the NetCDF file

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Extract the pressure, geopotential height, and wind variables

p = getvar(ncfile, "pressure")
z = getvar(ncfile, "z", units="dm")
ua = getvar(ncfile, "ua", units="kt")
va = getvar(ncfile, "va", units="kt")
wspd = getvar(ncfile, "wspd_wdir", units="kts")[0,:]

# Interpolate geopotential height, u, and v winds to 500 hPa

ht_500 = interplevel(z, p, 500)
u_500 = interplevel(ua, p, 500)
v_500 = interplevel(va, p, 500)
wspd_500 = interplevel(wspd, p, 500)

# Get the lat/lon coordinates

lats, lons = latlon_coords(ht_500)

# Get the basemap object

bm = get_basemap(ht_500)

# Create the figure

fig = plt.figure(figsize=(12,9))
ax = plt.axes()

# Convert the lat/lon coordinates to x/y coordinates in the projection space

x, y = bm(to_np(lons), to_np(lats))

# Add the 500 hPa geopotential height contours

levels = np.arange(520., 580., 6.)
contours = bm.contour(x, y, to_np(ht_500), levels=levels, colors="black")
plt.clabel(contours, inline=1, fontsize=10, fmt="%i")

# Add the wind speed contours

levels = [25, 30, 35, 40, 50, 60, 70, 80, 90, 100, 110, 120]
wspd_contours = bm.contourf(x, y, to_np(wspd_500), levels=levels,
cmap=get_cmap("rainbow"))
plt.colorbar(wspd_contours, ax=ax, orientation="horizontal", pad=.05)

# Add the geographic boundaries

bm.drawcoastlines(linewidth=0.25)
bm.drawstates(linewidth=0.25)
bm.drawcountries(linewidth=0.25)

# Add the 500 hPa wind barbs, only plotting every 125th data point.
bm.barbs(x[::125,::125], y[::125,::125], to_np(u_500[::125, ::125]),
to_np(v_500[::125, ::125]), length=6)

plt.title("500 MB Height (dm), Wind Speed (kt), Barbs (kt)")

plt.show()

56 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

Panel Plots from the Front Page

This lengthy example shows how to make the panel plots on the first page of the documentation. For a simpler example
of how to make a cross section plot, see Vertical Cross Section.

import numpy as np
import matplotlib.pyplot as plt
from matplotlib.cm import get_cmap
from netCDF4 import Dataset

from wrf import getvar, to_np, vertcross, smooth2d, CoordPair, get_basemap, latlon_
˓→coords

# Open the NetCDF file

ncfile = Dataset("wrfout_d01_2016-10-07_00_00_00")

# Get the WRF variables

slp = getvar(ncfile, "slp")
smooth_slp = smooth2d(slp, 3)
ctt = getvar(ncfile, "ctt")
z = getvar(ncfile, "z")
dbz = getvar(ncfile, "dbz")
(continues on next page)

1.5. Plotting Examples 57

wrf-python Documentation, Release 1.2.0

(continued from previous page)

Z = 10**(dbz/10.)
wspd = getvar(ncfile, "wspd_wdir", units="kt")[0,:]

# Set the start point and end point for the cross section
start_point = CoordPair(lat=26.76, lon=-80.0)
end_point = CoordPair(lat=26.76, lon=-77.8)

# Compute the vertical cross-section interpolation. Also, include the lat/lon points
# along the cross-section in the metadata by setting latlon to True.
z_cross = vertcross(Z, z, wrfin=ncfile, start_point=start_point, end_point=end_point,
latlon=True, meta=True)
wspd_cross = vertcross(wspd, z, wrfin=ncfile, start_point=start_point, end_point=end_
˓→point,

latlon=True, meta=True)
dbz_cross = 10.0 * np.log10(z_cross)

# Get the latitude and longitude points

lats, lons = latlon_coords(slp)

# Create the figure that will have 3 subplots

fig = plt.figure(figsize=(10,7))
ax_ctt = fig.add_subplot(1,2,1)
ax_wspd = fig.add_subplot(2,2,2)
ax_dbz = fig.add_subplot(2,2,4)

# Get the basemap object

bm = get_basemap(slp)

# Convert the lat/lon points in to x/y points in the projection space

x, y = bm(to_np(lons), to_np(lats))

# Make the pressure contours

contour_levels = [960, 965, 970, 975, 980, 990]
c1 = bm.contour(x, y, to_np(smooth_slp), levels=contour_levels, colors="white",
zorder=3, linewidths=1.0, ax=ax_ctt)

# Create the filled cloud top temperature contours

contour_levels = [-80.0, -70.0, -60, -50, -40, -30, -20, -10, 0, 10]
ctt_contours = bm.contourf(x, y, to_np(ctt), contour_levels, cmap=get_cmap("Greys"),
zorder=2, ax=ax_ctt)

point_x, point_y = bm([start_point.lon, end_point.lon], [start_point.lat, end_point.

˓→lat])

bm.plot([point_x[0], point_x[1]], [point_y[0], point_y[1]], color="yellow",

marker="o", zorder=3, ax=ax_ctt)

# Create the color bar for cloud top temperature

cb_ctt = fig.colorbar(ctt_contours, ax=ax_ctt, shrink=.60)
cb_ctt.ax.tick_params(labelsize=5)

# Draw the oceans, land, and states

bm.drawcoastlines(linewidth=0.25, ax=ax_ctt)
bm.drawstates(linewidth=0.25, ax=ax_ctt)
bm.drawcountries(linewidth=0.25, ax=ax_ctt)
bm.fillcontinents(color=np.array([ 0.9375 , 0.9375 , 0.859375]),
ax=ax_ctt, lake_color=np.array([ 0.59375 , 0.71484375, 0.8828125 ]))
bm.drawmapboundary(fill_color=np.array([ 0.59375 , 0.71484375, 0.8828125 ]), ax=ax_
˓→ctt) (continues on next page)

58 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

(continued from previous page)

# Draw Parallels
parallels = np.arange(np.amin(lats), 30., 2.5)
bm.drawparallels(parallels, ax=ax_ctt, color="white")

merids = np.arange(-85.0, -72.0, 2.5)

bm.drawmeridians(merids, ax=ax_ctt, color="white")

# Crop the image to the hurricane region

x_start, y_start = bm(-85.0, np.amin(lats))
x_end, y_end = bm(-72.0, 30.0)

ax_ctt.set_xlim([x_start, x_end])
ax_ctt.set_ylim([y_start, y_end])

# Make the contour plot for wspd

wspd_contours = ax_wspd.contourf(to_np(wspd_cross), cmap=get_cmap("jet"))
# Add the color bar
cb_wspd = fig.colorbar(wspd_contours, ax=ax_wspd)
cb_wspd.ax.tick_params(labelsize=5)

# Make the contour plot for dbz

# Set the x-ticks to use latitude and longitude labels.

# Set the y-ticks to be height.

# Set the x-axis and y-axis labels

ax_dbz.set_xlabel("Latitude, Longitude", fontsize=5)
ax_wspd.set_ylabel("Height (m)", fontsize=5)
ax_dbz.set_ylabel("Height (m)", fontsize=5)

# Add titles
ax_ctt.set_title("Cloud Top Temperature (degC)", {"fontsize" : 7})
ax_wspd.set_title("Cross-Section of Wind Speed (kt)", {"fontsize" : 7})
ax_dbz.set_title("Cross-Section of Reflectivity (dBZ)", {"fontsize" : 7})

plt.show()

1.5. Plotting Examples 59

wrf-python Documentation, Release 1.2.0

1.5.3 Vertical Cross Section

Vertical cross sections require no mapping software package and can be plotted using the standard matplotlib API.

import numpy as np
import matplotlib.pyplot as plt
from matplotlib.cm import get_cmap
import cartopy.crs as crs
from cartopy.feature import NaturalEarthFeature
from netCDF4 import Dataset

from wrf import to_np, getvar, CoordPair, vertcross

# Open the NetCDF file

filename = "wrfout_d01_2016-10-07_00_00_00"
ncfile = Dataset(filename)

# Extract the model height and wind speed

z = getvar(ncfile, "z")
wspd = getvar(ncfile, "uvmet_wspd_wdir", units="kt")[0,:]

# Create the start point and end point for the cross section
start_point = CoordPair(lat=26.76, lon=-80.0)
end_point = CoordPair(lat=26.76, lon=-77.8)

# Compute the vertical cross-section interpolation. Also, include the lat/lon

# points along the cross-section.
wspd_cross = vertcross(wspd, z, wrfin=ncfile, start_point=start_point, end_point=end_
˓→point,

latlon=True, meta=True)
(continues on next page)

60 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

(continued from previous page)

# Create the figure

fig = plt.figure(figsize=(12,6))
ax = plt.axes()

# Make the contour plot

wspd_contours = ax.contourf(to_np(wspd_cross), cmap=get_cmap("jet"))

# Add the color bar

plt.colorbar(wspd_contours, ax=ax)

# Set the x-ticks to use latitude and longitude labels.

coord_pairs = to_np(wspd_cross.coords["xy_loc"])
x_ticks = np.arange(coord_pairs.shape[0])
x_labels = [pair.latlon_str(fmt="{:.2f}, {:.2f}") for pair in to_np(coord_pairs)]
ax.set_xticks(x_ticks[::20])
ax.set_xticklabels(x_labels[::20], rotation=45, fontsize=8)

# Set the y-ticks to be height.

vert_vals = to_np(wspd_cross.coords["vertical"])
v_ticks = np.arange(vert_vals.shape[0])
ax.set_yticks(v_ticks[::20])
ax.set_yticklabels(vert_vals[::20], fontsize=8)

# Set the x-axis and y-axis labels

ax.set_xlabel("Latitude, Longitude", fontsize=12)
ax.set_ylabel("Height (m)", fontsize=12)

plt.title("Vertical Cross Section of Wind Speed (kt)")

plt.show()

1.6 API Reference

1.6.1 User API

Routines

Diagnostic Routine

The routine below is the primary routine for extracting variables from a WRF-ARW NetCDF file (or sequence of files)
and performing diagnostic calculations.

wrf.getvar Returns basic diagnostics from the WRF ARW model

output.

wrf.getvar

wrf.getvar(wrfin, varname, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

**kwargs)
Returns basic diagnostics from the WRF ARW model output.

1.6. API Reference 61

wrf-python Documentation, Release 1.2.0

A table of all available diagnostics is below.

Variable Name Description Available Units Additional Keyword

Arguments
avo Absolute Vorticity 10-5 s-1
eth/theta_e Equivalent Potential K units (str) : Set to de-
Temperature degC sired units. Default is
degF ‘K’.
cape_2d 2D cape J kg-1 ; J kg-1 ; m ; m missing (float): Fill
(mcape/mcin/lcl/lfc) value for output only
cape_3d 3D cape and cin J kg-1 missing (float): Fill
value for output only
ctt Cloud Top Temperature degC fill_nocloud (boolean):
K Set to True to use fill
degF values for cloud free re-
gions rather than surface
temperature. Default is
False.
missing (float): The
fill value to use when
fill_nocloud is True.
opt_thresh (float): The
optical depth required
to trigger the cloud top
temperature calculation.
Default is 1.0.
units (str) : Set to de-
sired units. Default is
‘degC’.
Continued on next page

62 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

Table 3 – continued from previous page

Variable Name Description Available Units Additional Keyword
Arguments
cloudfrac Cloud Fraction % vert_type (str): The
vertical coordinate type
for the cloud thresholds.
Must be ‘height_agl’,
‘height_msl’, or ‘pres’.
Default is ‘height_agl’.
low_thresh (float): The
low cloud threshold
(meters for ‘height_agl’
and ‘height_msl’, pas-
cals for ‘pres’). Default
is 300 m (97000 Pa)
mid_thresh (float): The
mid cloud threshold
(meters for ‘height_agl’
and ‘height_msl’, pas-
cals for ‘pres’). Default
is 2000 m (80000 Pa)
high_thresh (float): The
high cloud threshold
(meters for ‘height_agl’
and ‘height_msl’, pas-
cals for ‘pres’). Default
is 6000 m (45000 Pa)
dbz Reflectivity dBZ do_variant (boolean):
Set to True to enable
variant calculation.
Default is False.
do_liqskin (boolean):
Set to True to enable
liquid skin calculation.
Default is False.
mdbz Maximum Reflectivity dBZ do_variant (boolean):
Set to True to enable
variant calculation.
Default is False.
do_liqskin (boolean):
Set to True to enable
liquid skin calculation.
Default is False.
geopt/geopotential Geopotential for the m2 s-2
Mass Grid
geopt_stag Geopotential for the Ver- m2 s-2
tically Staggered Grid
helicity Storm Relative Helicity m2 s-2 top (float): The top
level for the calculation
in meters. Default is
3000.0.
lat Latitude decimal degrees
lon Longitude decimal degrees
Continued on next page

1.6. API Reference 63

wrf-python Documentation, Release 1.2.0

Table 3 – continued from previous page

Variable Name Description Available Units Additional Keyword
Arguments
omg/omega Omega Pa s-1
p/pres Full Model Pressure Pa units (str) : Set to de-
(in specified units) hPa sired units. Default is
mb ‘Pa’.
torr
mmhg
atm
pressure Full Model Pressure hPa
(hPa)
pvo Potential Vorticity PVU
pw Precipitable Water kg m-2
rh Relative Humidity %
rh2 2m Relative Humidity %
slp Sea Level Pressure hPa units (str) : Set to de-
hPa sired units. Default is
mb ‘hPa’.
torr
mmhg
atm
ter Model Terrain Height m units (str) : Set to de-
km sired units. Default is
dm ‘m’.
ft
mi
td2 2m Dew Point Tempera- degC units (str) : Set to de-
ture K sired units. Default is
degF ‘degC’.
td Dew Point Temperature degC units (str) : Set to de-
K sired units. Default is
degF ‘degC’.
tc Temperature in Celsius degC
th/theta Potential Temperature K units (str) : Set to de-
degC sired units. Default is
degF ‘K’.
temp Temperature (in speci- K units (str) : Set to de-
fied units) degC sired units. Default is
degF ‘K’.
tk Temperature in Kelvin K
times Times in the File or Se-
quence
xtimes XTIME Coordinate minutes since
(if applicable) start of
model run
tv Virtual Temperature K units (str) : Set to de-
degC sired units. Default is
degF ‘K’.
twb Wet Bulb Temperature K units (str) : Set to de-
degC sired units. Default is
degF ‘K’.
Continued on next page

64 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

Table 3 – continued from previous page

Variable Name Description Available Units Additional Keyword
Arguments
updraft_helicity Updraft Helicity m2 s-2 bottom (float): The bot-
tom level for the calcula-
tion in meters. Default is
2000.0.
top (float): The top
level for the calculation
in meters. Default is
5000.0.
ua U-component of Wind m s-1 units (str) : Set to de-
on Mass Points km h-1 sired units. Default is ‘m
mi h-1 s-1’.
kt
ft s-1
va V-component of Wind m s-1 units (str) : Set to de-
on Mass Points km h-1 sired units. Default is ‘m
mi h-1 s-1’.
kt
ft s-1
wa W-component of Wind m s-1 units (str) : Set to de-
on Mass Points km h-1 sired units. Default is ‘m
mi h-1 s-1’.
kt
ft s-1
uvmet10 10 m U and V Compo- m s-1 units (str) : Set to de-
nents of Wind km h-1 sired units. Default is ‘m
Rotated to Earth Coordi- mi h-1 s-1’.
nates kt
ft s-1
uvmet U and V Components of m s-1 units (str) : Set to de-
Wind km h-1 sired units. Default is ‘m
Rotated to Earth Coordi- mi h-1 s-1’.
nates kt
ft s-1
wspd_wdir Wind Speed m s-1 units (str) : Set to de-
and Direction km h-1 sired units. Default is ‘m
(wind_from_direction) mi h-1 s-1’.
in Grid Coordinates kt
ft s-1
wspd_wdir10 10m Wind Speed m s-1 units (str) : Set to de-
and Direction km h-1 sired units. Default is ‘m
(wind_from_direction) mi h-1 s-1’.
in Grid Coordinates kt
ft s-1
uvmet_wspd_wdir Wind Speed m s-1 units (str) : Set to de-
and Direction km h-1 sired units. Default is ‘m
(wind_from_direction) mi h-1 s-1’.
Rotated to Earth Coordi- kt
nates ft s-1
Continued on next page

1.6. API Reference 65

wrf-python Documentation, Release 1.2.0

Table 3 – continued from previous page

Variable Name Description Available Units Additional Keyword
Arguments
uvmet10_wspd_wdir 10m Wind Speed m s-1 units (str) : Set to de-
and Direction km h-1 sired units. Default is ‘m
(wind_from_direction) mi h-1 s-1’.
Rotated to Earth Coordi- kt
nates ft s-1
z/height Model Height for Mass m msl (boolean): Set to
Grid km False to return AGL val-
dm ues. True is for MSL.
ft Default is True.
mi units (str) : Set to de-
sired units. Default is
‘m’.
zstag Model Height for Verti- m msl (boolean): Set to
cally Staggered Grid km False to return AGL val-
dm ues. True is for MSL.
ft Default is True.
mi units (str) : Set to de-
sired units. Default is
‘m’.

66 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

• ValueError – Raised when an invalid diagnostic type or keyword argument is passed to

the routine.
• FortranError – Raised when a problem occurs during a Fortran calculation.

See also:
numpy.ndarray, xarray.DataArray

Examples
Using netCDF4

from netCDF4 import Dataset

from wrf import getvar

wrfnc = Dataset("wrfout_d02_2010-06-13_21:00:00")
slp = getvar(wrfnc, "slp")

Using PyNIO

from Nio import open_file

from wrf import getvar

wrfnc = open_file("wrfout_d02_2010-06-13_21:00:00"+".nc", "r")

slp = getvar(wrfnc, "slp")

Using Iterables:

import os
from netCDF4 import Dataset
from wrf import getvar

filedir = "/path/to/wrf/files"
wrfin = [Dataset(f) for f in os.listdir(filedir)
if f.startswith("wrfout_d02_")]

uvmet = getvar(wrfin, "uvmet", timeidx=3, units="kt")

Interpolation Routines

The routines below are the primary routines used for performing interpolation calculations.

wrf.interplevel Return the three-dimensional field interpolated to a hor-

izontal plane at the specified vertical level.
wrf.vertcross Return the vertical cross section for a three-dimensional
field.
wrf.interpline Return the two-dimensional field interpolated along a
line.
wrf.vinterp Return the field vertically interpolated to the given the
type of surface and a set of new levels.

1.6. API Reference 67

wrf-python Documentation, Release 1.2.0

wrf.interplevel

wrf.interplevel(field3d, vert, desiredlev, missing=9.969209968386869e+36, meta=True)

Return the three-dimensional field interpolated to a horizontal plane at the specified vertical level.
Parameters
• field3d (xarray.DataArray or numpy.ndarray) – A three-dimensional field to
interpolate, with the rightmost dimensions of nz x ny x nx.
• vert (xarray.DataArray or numpy.ndarray) – A three-dimensional array for the
vertical coordinate, typically pressure or height. This array must have the same dimension-
ality as field3d.
• desiredlev (float) – The desired vertical level. Must be in the same units as the vert
parameter.
• missing (float) – The fill value to use for the output. Default is wrf.
default_fill(numpy.float64).
• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.
Returns The interpolated variable. If xarray is enabled and the meta parameter is True, then the re-
sult will be an xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

Example
Interpolate Geopotential Height to 500 hPa

from netCDF4 import Dataset

from wrf import getvar, interplevel

wrfin = Dataset("wrfout_d02_2010-06-13_21:00:00")

p = getvar(wrfin, "pressure")
ht = getvar(wrfin, "z", units="dm")

ht_500 = interplevel(ht, p, 500.0)

wrf.vertcross

wrf.vertcross(field3d, vert, levels=None, missing=9.969209968386869e+36, wrfin=None, timeidx=0,

stagger=None, projection=None, ll_point=None, pivot_point=None, angle=None,
start_point=None, end_point=None, latlon=False, cache=None, meta=True)
Return the vertical cross section for a three-dimensional field.
The cross section is defined by a horizontal line through the domain. This horizontal line is defined by
either including the pivot_point and angle parameters, or the start_point and end_point parameters. The
pivot_point, start_point, and end_point coordinates can be defined in either x,y or latitude,longitude space.
If latitude,longitude coordinates are used, then a WRF input file or map projection must also be specified.
The vertical levels for the cross section are fixed if levels is not specified, and are determined by dividing
the vertical coordinate in to grid boxes of roughly 1% of the maximum vertical distance from top to bottom.

68 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

Otherwise, the levels argument can be used to specify specific vertical levels. If all vertical levels are desired,
use the raw wrf.interp2dxy() function.
See also:
wrf.interp2dxy()

Parameters
• field3d (xarray.DataArray or numpy.ndarray) – A three-dimensional field to
interpolate, whose rightmost dimensions are nz x ny x nx.
• vert (xarray.DataArray or numpy.ndarray) – A three-dimensional variable for
the vertical coordinate, typically pressure or height. This array must have the same dimen-
sionality as field3d
• levels (sequence, optional) – A sequence of float for the desired vertical levels
in the output array. Must be in the same units as vert. If None, a fixed set of vertical levels
is provided. Default is None.
• missing (float) – The fill value to use for the output. Default is wrf.
default_fill(numpy.float64).
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable, optional) – WRF-ARW
NetCDF data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of
the aforementioned types. This is used to obtain the map projection when using lati-
tude,longitude coordinates. Default is None.
• timeidx (int, optional) – The desired time index when obtaining map boundary informa-
tion from moving nests. This value can be a positive or negative integer. Only required when
wrfin is specified and the nest is moving. Currently, wrf.ALL_TIMES is not supported.
Default is 0.
• stagger (str) – If using latitude, longitude coordinate pairs for start_point, end_point,
or pivot_point, set the appropriate grid staggering type for field3d. By default, the mass grid
is used. The options are:
– ’m’: Use the mass grid (default).
– ’u’: Use the same staggered grid as the u wind component, which has a staggered
west_east (x) dimension.
– ’v’: Use the same staggered grid as the v wind component, which has a staggered
south_north (y) dimension.
• projection (wrf.WrfProj subclass, optional) – The map projection object to use
when working with latitude, longitude coordinates, and must be specified if wrfin is None.
Default is None.
• ll_point (wrf.CoordPair, sequence of wrf.CoordPair, optional) – The lower
left latitude, longitude point for your domain, and must be specified if wrfin is None. If the
domain is a moving nest, this should be a sequence of wrf.CoordPair. Default is None.
• pivot_point (wrf.CoordPair, optional) – A coordinate pair for the pivot point,
which indicates the location through which the plane will pass. Must also specify angle.
The coordinate pair can be in x,y grid coordinates or latitude, longitude coordinates. If us-
ing latitude, longitude coordinates, then either wrfin or projection must be specified to obtain
the map projection. Default is None.
• angle (float, optional) – Only valid for cross sections where a plane will be plotted
through a given point on the model domain. 0.0 represents a S-N cross section. 90.0 is a
W-E cross section.

1.6. API Reference 69

wrf-python Documentation, Release 1.2.0

• start_point (wrf.CoordPair, optional) – A coordinate pair which indicates the start

location through which the plane will pass. Must also specify end_point. The coordinate pair
can be in x,y grid coordinates or latitude, longitude coordinates. If using latitude, longitude
coordinates, then either wrfin or projection must be specified to obtain the map projection.
Default is None.
• end_point (wrf.CoordPair, optional) – A coordinate pair which indicates the end
location through which the plane will pass. Must also specify end_point. The coordinate
pair can be in x,y grid coordinates or latitude, longitude coordinates. If using latitude,
longitude coordinates, then either wrfin or projection must be specified to obtain the map
projection. Default is None.
• latlon (bool, optional) – Set to True to also interpolate the two-dimensional latitude
and longitude coordinates along the same horizontal line and include this information in the
metadata (if enabled). This can be helpful for plotting. Default is False.

Note: Currently, field3d must be of type xarray.DataArray and contain coordinate

information in order to generate the latitude and longitude coordinates along the line if
latlon is set to True. Otherwise, a warning will be issued, and the latitude and longitude
information will not be present.

• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
Returns The interpolated variable. If xarray is enabled and the meta parameter is True, then the re-
sult will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.interpline

wrf.interpline(field2d, pivot_point=None, wrfin=None, timeidx=0, stagger=None, projection=None,

ll_point=None, angle=None, start_point=None, end_point=None, latlon=False,
cache=None, meta=True)
Return the two-dimensional field interpolated along a line.
This line is defined by either including the pivot_point and angle parameters, or the start_point and end_point
parameters. The pivot_point, start_point, and end_point coordinates can be defined in either x,y or lati-
tude,longitude space. If latitude,longitude coordinates are used, then a WRF input file or map projection must
also be specified.
Parameters
• field2d (xarray.DataArray or numpy.ndarray) – A two-dimensional field.
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable, optional) – WRF-ARW
NetCDF data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of
the aforementioned types. This is used to obtain the map projection when using lati-
tude,longitude coordinates. Should not be used when working with x,y coordinates. Default
is None.

70 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

• timeidx (int, optional) – The desired time index when obtaining map boundary informa-
tion from moving nests. This value can be a positive or negative integer. Only required when
wrfin is specified and the nest is moving. Currently, wrf.ALL_TIMES is not supported.
Default is 0.
• stagger (str) – If using latitude, longitude coordinate pairs for start_point, end_point,
or pivot_point, set the appropriate grid staggering type for field2d. By default, the mass grid
is used. The options are:
– ’m’: Use the mass grid (default).
– ’u’: Use the same staggered grid as the u wind component, which has a staggered
west_east (x) dimension.
– ’v’: Use the same staggered grid as the v wind component, which has a staggered
south_north (y) dimension.
• projection (wrf.WrfProj, optional) – The map projection object to use when work-
ing with latitude, longitude coordinates, and must be specified if wrfin is None. Should not
be used when working with x,y coordinates. Default is None.
• ll_point (wrf.CoordPair, sequence of wrf.CoordPair, optional) – The lower
left latitude, longitude point for your domain, and must be specified if wrfin is None. If the
domain is a moving nest, this should be a sequence of wrf.CoordPair. Default is None.
• pivot_point (wrf.CoordPair, optional) – A coordinate pair for the pivot point,
which indicates the location through which the plane will pass. Must also specify angle.
The coordinate pair can be in x,y grid coordinates or latitude, longitude coordinates. If us-
ing latitude, longitude coordinates, then either wrfin or projection must be specified to obtain
the map projection. Default is None.
• angle (float, optional) – Only valid for cross sections where a plane will be plotted
through a given point on the model domain. 0.0 represents a S-N cross section. 90.0 is a
W-E cross section.
• start_point (wrf.CoordPair, optional) – A coordinate pair which indicates the start
location through which the plane will pass. Must also specify end_point. The coordinate pair
can be in x,y grid coordinates or latitude, longitude coordinates. If using latitude, longitude
coordinates, then either wrfin or projection must be specified to obtain the map projection.
Default is None.
• end_point (wrf.CoordPair, optional) – A coordinate pair which indicates the end
location through which the plane will pass. Must also specify end_point. The coordinate
pair can be in x,y grid coordinates or latitude, longitude coordinates. If using latitude,
longitude coordinates, then either wrfin or projection must be specified to obtain the map
projection. Default is None.
• latlon (bool, optional) – Set to True to also interpolate the two-dimensional latitude
and longitude coordinates along the same horizontal line and include this information in the
metadata (if enabled). This can be helpful for plotting. Default is False.

Note: Currently, field2d must be of type xarray.DataArray and contain coordinate

• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-

1.6. API Reference 71

wrf-python Documentation, Release 1.2.0

ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
Returns The interpolated variable. If xarray is enabled and the meta parameter is True, then the re-
sult will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.vinterp

wrf.vinterp(wrfin, field, vert_coord, interp_levels, extrapolate=False, field_type=None, log_p=False,

timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True)
Return the field vertically interpolated to the given the type of surface and a set of new levels.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• field (xarray.DataArray or numpy.ndarray) – A three-dimensional field.
• vert_coord (str) – A string indicating the vertical coordinate type to interpolate to.
Valid strings are:
– ’pressure’, ‘pres’, ‘p’: pressure [hPa]
– ’ght_msl’: grid point height msl [km]
– ’ght_agl’: grid point height agl [km]
– ’theta’, ‘th’: potential temperature [K]
– ’theta-e’, ‘thetae’, ‘eth’: equivalent potential temperature [K]
• interp_levels (sequence) – A 1D sequence of vertical levels to interpolate to.
• extrapolate (bool, optional) – Set to True to extrapolate values below ground. Default
is False.
• field_type (str, optional) – The type of field. Default is None.
Valid strings are:
– ’none’: None
– ’pressure’, ‘pres’, ‘p’: pressure
– ’z’, ‘ght’: geopotential height
– ’tc’: temperature [degC]
– ’tk’: temperature [K]
– ’theta’, ‘th’: potential temperature [K]
– ’theta-e’, ‘thetae’, ‘eth’: equivalent potential temperature
• log_p (bool, optional) – Use the log of the pressure for interpolation instead of pressure.
Default is False.

72 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

• timeidx (int, optional) – The time index to use when extracting auxiallary variables
used in the interpolation. This value must be set to match the same value used when the
field variable was extracted. Default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
Returns The interpolated variable. If xarray is enabled and the meta parameter is True, then the re-
sult will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

Lat-Lon to/from XY Routines

The routines below are used for converting back and forth between xy-grid space and latitude-longitude space.

wrf.ll_to_xy Return the x,y coordinates for a specified latitude and

longitude.
wrf.xy_to_ll Return the latitude and longitude for specified x,y coor-
dinates.
wrf.ll_to_xy_proj Return the x, y coordinates for a specified latitude and
longitude.
wrf.xy_to_ll_proj Return the latitude and longitude for the specified x,y
coordinates.

wrf.ll_to_xy

wrf.ll_to_xy(wrfin, latitude, longitude, timeidx=0, squeeze=True, meta=True, stagger=None,

as_int=True)
Return the x,y coordinates for a specified latitude and longitude.
The latitude and longitude arguments can be a single value or a sequence of values.
The leftmost dimension of the returned array represents two different quantities:
• return_val[0,. . . ] will contain the X (west_east) values.
• return_val[1,. . . ] will contain the Y (south_north) values.

Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF

1.6. API Reference 73

wrf-python Documentation, Release 1.2.0

data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-

tioned types.
• latitude (float or sequence) – A single latitude or a sequence of latitude values to be
converted.
• longitude (float or sequence) – A single longitude or a sequence of latitude values to
be converted.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• stagger (str) – By default, the latitude is returned on the mass grid, but a staggered grid
can be chosen with the following options:
– ’m’: Use the mass grid (default).
– ’u’: Use the same staggered grid as the u wind component, which has a staggered
west_east (x) dimension.
– ’v’: Use the same staggered grid as the v wind component, which has a staggered
south_north (y) dimension.
• as_int (bool) – Set to True to return the x,y values as int, otherwise they will be
returned as float.
Returns The x,y coordinate value(s) whose leftmost dimension is 2 (0=X, 1=Y). If xarray is en-
abled and the meta parameter is True, then the result will be a xarray.DataArray object.
Otherwise, the result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.xy_to_ll

wrf.xy_to_ll(wrfin, x, y, timeidx=0, squeeze=True, meta=True, stagger=None)

Return the latitude and longitude for specified x,y coordinates.
The x and y arguments can be a single value or a sequence of values.
The leftmost dimension of the returned array represents two different quantities:
• return_val[0,. . . ] will contain the latitude values.
• return_val[1,. . . ] will contain the longitude values.

74 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

• y (float or sequence) – A single y-coordinate or a sequence of y-coordinate values to be

converted.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• stagger (str) – By default, the latitude is returned on the mass grid, but a staggered grid
can be chosen with the following options:
– ’m’: Use the mass grid (default).
– ’u’: Use the same staggered grid as the u wind component, which has a staggered
west_east (x) dimension.
– ’v’: Use the same staggered grid as the v wind component, which has a staggered
south_north (y) dimension.
Returns The latitude and longitude values whose leftmost dimension is 2 (0=latitude, 1=longi-
tude). If xarray is enabled and the meta parameter is True, then the result will be a xarray.
DataArray object. Otherwise, the result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.ll_to_xy_proj

wrf.ll_to_xy_proj(latitude, longitude, meta=True, squeeze=True, as_int=True, map_proj=None,

truelat1=None, truelat2=None, stand_lon=None, ref_lat=None, ref_lon=None,
pole_lat=None, pole_lon=None, known_x=None, known_y=None, dx=None,
dy=None, latinc=None, loninc=None)
Return the x, y coordinates for a specified latitude and longitude.
The latitude and longitude arguments can be a single value or a sequence of values. This version of the ll_to_xy
routine allows users to manually specify projection parameters.
The leftmost dimension of the returned array represents two different quantities:
• return_val[0,. . . ] will contain the X (west_east) values.
• return_val[1,. . . ] will contain the Y (south_north) values.

Parameters
• latitude (float or sequence) – A single latitude or a sequence of latitude values to be
converted.
• longitude (float or sequence) – A single longitude or a sequence of latitude values to
be converted.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.

1.6. API Reference 75

wrf-python Documentation, Release 1.2.0

• as_int (bool) – Set to True to return the x,y values as int, otherwise they will be
returned as float.
• map_proj (int) – Model projection [1=Lambert Conformal, 2=Polar Stereographic,
3=Mercator, 6=Lat-Lon]. Required.
• truelat1 (float) – True latitude 1. Required for map_proj = 1, 2, 3 (defaults to 0
otherwise).
• truelat2 (float) – True latitude 2. Optional for map_proj = 1 (defaults to 0 otherwise).
• stand_lon (float) – Standard longitude. Required.
• ref_lat (float) – A reference latitude. Required.
• ref_lon (float) – A reference longitude. Required.
• known_x (float) – The known x-coordinate associated with ref_lon. Required.
• known_y (float) – The known y-coordinate associated with ref_lat. Required.
• pole_lat (float) – Pole latitude. Optional for map_proj = 6 (defaults to 90 otherwise).
• pole_lon (float) – Pole longitude. Optional for map_proj = 6 (defaults to 0 otherwise).
• dx (float) – The x spacing in meters at the true latitude. Required for map_proj = 1, 2, 3
(defaults to 0 otherwise).
• dy (float) – Required for map_proj = 1, 2, 3 (defaults to 0 otherwise).
• latinc (float) – Required for map_proj = 6. Defined as:

latinc = (dy*360.0)/2.0/Constants.PI/Constants.WRF_EARTH_RADIUS

• loninc (float) – Required for map_proj = 6. Defined as:

loninc = (dx*360.0)/2.0/Constants.PI/Constants.WRF_EARTH_RADIUS

Returns The x,y coordinate value(s) whose leftmost dimension is 2 (0=X, 1=Y). If xarray is en-
abled and the meta parameter is True, then the result will be a xarray.DataArray object.
Otherwise, the result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.xy_to_ll_proj

wrf.xy_to_ll_proj(x, y, meta=True, squeeze=True, map_proj=None, truelat1=None, truelat2=None,

stand_lon=None, ref_lat=None, ref_lon=None, pole_lat=None, pole_lon=None,
known_x=None, known_y=None, dx=None, dy=None, latinc=None, loninc=None)
Return the latitude and longitude for the specified x,y coordinates.
The x and y arguments can be a single value or a sequence of values. This version of the xy_to_ll routine allows
users to manually specify map projection parameters.
The leftmost dimension of the returned array represents two different quantities:
• return_val[0,. . . ] will contain the latitude values.
• return_val[1,. . . ] will contain the longitude values.

Parameters

76 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

• x (float or sequence) – A single x-coordinate or a sequence of x-coordinate values to be

converted.
• y (float or sequence) – A single y-coordinate or a sequence of y-coordinate values to be
converted.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• as_int (bool) – Set to True to return the x,y values as int, otherwise they will be
returned as float.
• map_proj (int) – Model projection [1=Lambert Conformal, 2=Polar Stereographic,
3=Mercator, 6=Lat-Lon]. Required.
• truelat1 (float) – True latitude 1. Required for map_proj = 1, 2, 3 (defaults to 0
otherwise).
• truelat2 (float) – True latitude 2. Optional for map_proj = 1 (defaults to 0 otherwise).
• stand_lon (float) – Standard longitude. Required.
• ref_lat (float) – A reference latitude. Required.
• ref_lon (float) – A reference longitude. Required.
• known_x (float) – The known x-coordinate associated with ref_lon. Required.
• known_y (float) – The known y-coordinate associated with ref_lat. Required.
• pole_lat (float) – Pole latitude. Optional for map_proj = 6 (defaults to 90 otherwise).
• pole_lon (float) – Pole longitude. Optional for map_proj = 6 (defaults to 0 otherwise).
• dx (float) – The x spacing in meters at the true latitude. Required for map_proj = 1, 2, 3
(defaults to 0 otherwise).
• dy (float) – Required for map_proj = 1, 2, 3 (defaults to 0 otherwise).
• latinc (float) – Required for map_proj = 6. Defined as:

latinc = (dy*360.0)/2.0/Constants.PI/Constants.WRF_EARTH_RADIUS

• loninc (float) – Required for map_proj = 6. Defined as:

loninc = (dx*360.0)/2.0/Constants.PI/Constants.WRF_EARTH_RADIUS

Returns The latitude and longitude values whose leftmost dimension is 2 (0=latitude, 1=longi-
tude). If xarray is enabled and the meta parameter is True, then the result will be a xarray.
DataArray object. Otherwise, the result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

Grid Destaggering Routine

The routine below is used to convert a variable on a staggered grid to the unstaggered grid.

wrf.destagger Return the variable on the unstaggered grid.

1.6. API Reference 77

wrf-python Documentation, Release 1.2.0

wrf.destagger

wrf.destagger(var, stagger_dim, meta=False)

Return the variable on the unstaggered grid.
This function destaggers the variable by taking the average of the values located on either side of the grid box.
Parameters
• var (xarray.DataArray or numpy.ndarray) – A variable on a staggered grid.
• stagger_dim (int) – The dimension index to destagger. Negative values can be used to
choose dimensions referenced from the right hand side (-1 is the rightmost dimension).
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is False.
Returns The destaggered variable. If xarray is enabled and the meta parameter is True, then the re-
sult will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

Numpy Extraction Routine

The routine below is used to extract a numpy.ndarray from a xarray.DataArray. This routine must be used
before passing the array object to a compiled extension.

wrf.to_np Return the numpy.ndarray contained in an

xarray.DataArray instance.

wrf.to_np

wrf.to_np(array)
Return the numpy.ndarray contained in an xarray.DataArray instance.
If the xarray.DataArray instance does not contain a _FillValue or missing_value attribute, then this routine
simply returns the xarray.DataArray.values attribute. If the xarray.DataArray object contains a
_FillValue or missing_value attribute, then this routine returns a numpy.ma.MaskedArray instance, where
the NaN values (used by xarray to represent missing data) are replaced with the fill value.
If the object passed in to this routine is not an xarray.DataArray instance, then this routine simply returns
the passed in object. This is useful in situations where you do not know if you have an xarray.DataArray
or a numpy.ndarray and simply want a numpy.ndarray returned.
Parameters array (xarray.DataArray, numpy.ndarray, or any object) – Can be any ob-
ject type, but is generally used with xarray.DataArray or numpy.ndarray.
Returns The extracted array or the array object if array is not a xarray.DataArray object..
Return type numpy.ndarray or numpy.ma.MaskedArray

Variable Extraction Routines

The routines below are primarily used internally by wrf.getvar(), but some users may find them useful to manu-
ally extract variables from a WRF NetCDF file (or a sequence of NetCDF files).

78 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

wrf.extract_vars Extract variables from a NetCDF file object or a se-

quence of NetCDF file objects.
wrf.combine_files Combine and return an array object for the sequence of
WRF output files.
wrf.extract_dim Return the dimension size for the specified dimension
name.
wrf.extract_global_attrs Return the global attribute(s).
wrf.extract_times Return a sequence of time objects.

wrf.extract_vars

wrf.extract_vars(wrfin, timeidx, varnames, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None)
Extract variables from a NetCDF file object or a sequence of NetCDF file objects.
Parameters
• wrfin (iterable) – An iterable type, which includes lists, tuples, dictionaries, genera-
tors, and user-defined classes.
• varnames (sequence of str) – A sequence of variable names.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – Cache key for the coordinate variables. This is used for internal
purposes only. Default is None.
Returns A mapping of variable name to an array object. If xarray is enabled and the meta parameter
is True, then the array object will be a xarray.DataArray object. Otherwise, the array
object will be a numpy.ndarray object with no metadata.
Return type dict

wrf.combine_files

wrf.combine_files(wrfin, varname, timeidx, is_moving=None, method=’cat’, squeeze=True,

meta=True, _key=None)
Combine and return an array object for the sequence of WRF output files.

1.6. API Reference 79

wrf-python Documentation, Release 1.2.0

Two aggregation methodologies are available to combine the sequence:

• ‘cat’: Concatenate the files along the ‘Time’ dimension. The Time dimension will be the leftmost di-
mension. No sorting is performed, so files must be properly ordered in the sequence prior to calling this
function.
• ‘join’: Join the files by creating a new leftmost dimension for the file index. In situations where there are
multiple files with multiple times, and the last file contains less times than the previous files, the remaining
arrays will be arrays filled with missing values. There are checks in place within the wrf-python algorithms
to look for these missing arrays, but be careful when calling compiled routines outside of wrf-python.

Parameters
• wrfin (iterable) – An iterable type, which includes lists, tuples, dictionaries, genera-
tors, and user-defined classes.
• varname (str) – The variable name.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• is_moving (bool) – A boolean type that indicates if the sequence is a moving nest.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – Cache key for the coordinate variables. This is used for internal
purposes only. Default is None.
Returns If xarray is enabled and the meta parameter is True, then the result will be a xarray.
DataArray object. Otherwise, the result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.extract_dim

wrf.extract_dim(wrfin, dim)
Return the dimension size for the specified dimension name.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• dim (str) – The dimension name.
Returns The dimension size.
Return type int

80 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

wrf.extract_global_attrs

wrf.extract_global_attrs(wrfin, attrs)
Return the global attribute(s).
If the wrfin parameter is a sequence, then only the first element is used, so the entire sequence must have the
same global attributes.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• attrs (str, sequence) – The attribute name or a sequence of attribute names.
Returns A mapping of attribute_name to value.
Return type dict

wrf.extract_times

wrf.extract_times(wrfin, timeidx, method=’cat’, squeeze=True, cache=None, meta=False,

do_xtime=False)
Return a sequence of time objects.
If do_xtime is False, the ‘XTIME’ variable is used and each time object is a float. Otherwise, the ‘Times’
variable is used, and each time object is a datetime.datetime object.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES) – The desired time index. This value can be a
positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata.
• do_xtime (bool) – Set to True to parse the ‘XTIME’ variable instead of the ‘Times’
variable. Default is False.
Returns A sequence of time objects. If meta is True, the sequence will be of type xarray.
DataArray, otherwise the sequence is numpy.ndarray.
Return type xarray.DataArray or numpy.ndarray

1.6. API Reference 81

wrf-python Documentation, Release 1.2.0

Plotting Helper Routines

The routines below are used to assist with plotting.

wrf.geo_bounds Return the geographic boundaries for the variable or

file(s).
wrf.latlon_coords Return the latitude and longitude coordinates from a
xarray.DataArray object.
wrf.get_cartopy Return a cartopy.crs.Projection subclass for
the map projection.
wrf.get_basemap Return a matplotlib.mpl_toolkits.
basemap.Basemap object
wrf.get_pyngl Return a Ngl.Resources object for the map projec-
tion.
wrf.cartopy_xlim Return the x-axis limits in the projected coordinates.
wrf.cartopy_ylim Return the y-axis limits in the projected coordinates.

wrf.geo_bounds

wrf.geo_bounds(var=None, wrfin=None, varname=None, timeidx=0, method=’cat’, squeeze=True,

cache=None)
Return the geographic boundaries for the variable or file(s).
When using a xarray.DataArray as the var parameter, the variable must contain latitude and longitude co-
ordinates. If these coordinate dimensions are greater than two dimensions, then an array of wrf.GeoBounds
objects will be returned with the same shape as the leftmost dimensions of the coordinate arrays.
When using a WRF file, or sequence of WRF files, by supplying the wrfin parameter, an array of wrf.
GeoBounds objects will be returned if the domain is moving and wrf.ALL_TIMES is selected as the timeidx
parameter when using wrfin. Otherwise, a single wrf.GeoBounds object is returned.
Parameters
• var (xarray.DataArray, optional) – A xarray.DataArray variable that includes
latitude,longitude coordinate information. If not used, then wrfin must be provided.
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable, optional) – WRF-ARW
NetCDF data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the
aforementioned types. If not used, then var must be provided.
• varname (str, optional) – If using wrfin, then this will be the variable name to use to
determine the geobounds. The variable can be a coordinate variable, or a regular vari-
able that contains coordinate attributes. If None, then the ‘XLAT’, ‘XLAT_M’, ‘XLONG’,
‘XLONG_M’ variables will be used.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index when wrfin is
not None. This value can be a positive integer, negative integer, or wrf.ALL_TIMES (an
alias for None) to return all times in the file or sequence. Default is 0. This value is ignored
when var is used.
• method (str, optional) – The aggregation method to use for sequences when wrfin is not
None. Must be either ‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension.
‘join’ creates a new dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Only used when wrfin is used. Default
is True.

82 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

wrf.latlon_coords

wrf.latlon_coords(var, as_np=False)
Return the latitude and longitude coordinates from a xarray.DataArray object.
Parameters
• var (xarray.DataArray) – A variable.
• to_np (bool) – Set to True to return the coordinates as numpy.ndarray objects instead
of xarray.DataArray objects.
Returns The latitude and longitude coordinate variables.
Return type tuple

wrf.get_cartopy

wrf.get_cartopy(var=None, wrfin=None, varname=None, timeidx=0, method=’cat’, squeeze=True,

cache=None)
Return a cartopy.crs.Projection subclass for the map projection.
Parameters
• var (xarray.DataArray, optional) – A xarray.DataArray variable that includes
latitude,longitude coordinate information. If not used, then wrfin must be provided.
• geobounds (wrf.GeoBounds, optional) – The geobounds to get the extents. If set to
None and using the var parameter, the geobounds will be taken from the variable. If using a
file, then the geobounds will be taken from the native grid.
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable, optional) – WRF-ARW
NetCDF data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the
aforementioned types. If not used, then var must be provided.
• varname (str, optional) – If using wrfin, then this will be the variable name to use to
determine the geobounds. The variable can be a coordinate variable, or a regular vari-
able that contains coordinate attributes. If None, then the ‘XLAT’, ‘XLAT_M’, ‘XLONG’,
‘XLONG_M’ variables will be used.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. Default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.

1.6. API Reference 83

wrf-python Documentation, Release 1.2.0

wrf.get_basemap

wrf.get_basemap(var=None, wrfin=None, varname=None, timeidx=0, method=’cat’, squeeze=True,

cache=None, **kwargs)
Return a matplotlib.mpl_toolkits.basemap.Basemap object for the map projection.

Parameters
• var (xarray.DataArray, optional) – A xarray.DataArray variable that includes
latitude,longitude coordinate information. If not used, then wrfin must be provided.
• geobounds (wrf.GeoBounds, optional) – The geobounds to get the extents. If set to
None and using the var parameter, the geobounds will be taken from the variable. If using a
file, then the geobounds will be taken from the native grid.
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable, optional) – WRF-ARW
NetCDF data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the
aforementioned types. If not used, then var must be provided.
• varname (str, optional) – If using wrfin, then this will be the variable name to use to
determine the geobounds. The variable can be a coordinate variable, or a regular vari-
able that contains coordinate attributes. If None, then the ‘XLAT’, ‘XLAT_M’, ‘XLONG’,
‘XLONG_M’ variables will be used.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. Default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• **kwargs – Keyword arguments for creating a matplotlib.mpl_toolkits.
basemap.Basemap. By default, the domain bounds will be set to the native projection,
the resolution will be set to ‘l’, and the other projection parameters will be set by the infor-
mation in the file.

84 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

Returns A Projection subclass for the map projection.

Return type cartopy.crs.Projection
Returns A Basemap object for the projection.
Return type
matplotlib.mpl_toolkits.basemap.Basemap
See Also:
matplotlib.mpl_toolkits.basemap.Basemap

wrf.get_pyngl

wrf.get_pyngl(var=None, wrfin=None, varname=None, timeidx=0, method=’cat’, squeeze=True,

cache=None, **kwargs)
Return a Ngl.Resources object for the map projection.
Parameters
• var (xarray.DataArray, optional) – A xarray.DataArray variable that includes
latitude,longitude coordinate information. If not used, then wrfin must be provided.
• geobounds (wrf.GeoBounds, optional) – The geobounds to get the extents. If set to
None and using the var parameter, the geobounds will be taken from the variable. If using a
file, then the geobounds will be taken from the native grid.
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable, optional) – WRF-ARW
NetCDF data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the
aforementioned types. If not used, then var must be provided.
• varname (str, optional) – If using wrfin, then this will be the variable name to use to
determine the geobounds. The variable can be a coordinate variable, or a regular vari-
able that contains coordinate attributes. If None, then the ‘XLAT’, ‘XLAT_M’, ‘XLONG’,
‘XLONG_M’ variables will be used.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. Default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• **kwargs – Additional PyNGL resources to set while creating the Ngl.Resources
object.
Returns A dict-like object that contains the PyNGL resources for the map projection.
Return type Ngl.Resources

1.6. API Reference 85

wrf-python Documentation, Release 1.2.0

wrf.cartopy_xlim(var=None, geobounds=None, wrfin=None, varname=None, timeidx=0,

method=’cat’, squeeze=True, cache=None)
Return the x-axis limits in the projected coordinates.
For some map projections, like :class‘wrf.RotatedLatLon‘, the cartopy.GeoAxes.set_extent()
method does not work correctly. This method is equivalent to:

pc = crs.PlateCarree()
xs, ys, _ = self._cartopy().transform_points(pc,
np.array([geobounds.bottom_left.lon,
geobounds.top_right.lon]),
np.array([geobounds.bottom_left.lat,
geobounds.top_right.lat])).T

_xlimits = xs.tolist()
_ylimits = ys.tolist()

return (_xlimits, _ylimits)[0]

86 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
Returns A list of [start_x, end_x] in the projected coordinate system.
Return type list

wrf.cartopy_ylim

wrf.cartopy_ylim(var=None, geobounds=None, wrfin=None, varname=None, timeidx=0,

method=’cat’, squeeze=True, cache=None)
Return the y-axis limits in the projected coordinates.
For some map projections, like :class‘wrf.RotatedLatLon‘, the cartopy.GeoAxes.set_extent()
method does not work correctly. This method is equivalent to:

_xlimits = xs.tolist()
_ylimits = ys.tolist()

return (_xlimits, _ylimits)[1]

1.6. API Reference 87

wrf-python Documentation, Release 1.2.0

Raw Diagnostic Routines

The routines below can be used when working with variables that are not contained in a WRF-ARW NetCDF file.
They can also be used with non-WRF data. However, if you are working with WRF-ARW NetCDF files, use wrf.
getvar() instead.
Keep in mind that these routines were developed for WRF-ARW, so your mileage may vary when working with non-
WRF data. Also, the vast majority of these routines do not allow for missing values in any of the input arrays, so make
sure they are removed before calling these routines.

wrf.xy Return the x,y points for a line within a two-dimensional

grid.
wrf.interp1d Return the linear interpolation of a one-dimensional
variable.
wrf.interp2dxy Return a cross section for a three-dimensional field.
wrf.interpz3d Return the field interpolated to a specified pressure or
height level.
wrf.slp Return the sea level pressure.
wrf.tk Return the temperature.
wrf.td Return the dewpoint temperature.
wrf.rh Return the relative humidity.
wrf.uvmet Return the u,v components of the wind rotated to earth
coordinates.
wrf.smooth2d Return the field smoothed.
wrf.cape_2d Return the two-dimensional MCAPE, MCIN, LCL, and
LFC.
wrf.cape_3d Return the three-dimensional CAPE and CIN.
wrf.cloudfrac Return the cloud fraction.
wrf.ctt Return the cloud top temperature.
wrf.dbz Return the simulated radar reflectivity.
wrf.srhel Return the storm relative helicity.
wrf.udhel Return the updraft helicity.
wrf.avo Return the absolute vorticity.
wrf.pvo Return the potential vorticity.
wrf.eth Return the equivalent potential temperature.
wrf.wetbulb Return the wetbulb temperature.
wrf.tvirtual Return the virtual temperature.
wrf.omega Return omega.
wrf.pw Return the precipitable water.

88 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

wrf.xy

wrf.xy(field, pivot_point=None, angle=None, start_point=None, end_point=None, meta=True)

Return the x,y points for a line within a two-dimensional grid.
This function is primarily used to obtain the x,y points when making a cross section.
Parameters
• field (xarray.DataArray or numpy.ndarray) – A field with at least two dimen-
sions.
• pivot_point (tuple or list, optional) – A tuple or list with two entries, in the
form of [x, y] (or [west_east, south_north]), which indicates the x,y location through which
the plane will pass. Must also specify angle.
• angle (float, optional) – Only valid for cross sections where a plane will be plotted
through a given point on the model domain. 0.0 represents a S-N cross section. 90.0 is a
W-E cross section.
• start_point (tuple or list, optional) – A tuple or list with two entries, in the
form of [x, y] (or [west_east, south_north]), which indicates the start x,y location through
which the plane will pass.
• end_point (tuple or list, optional) – A tuple or list with two entries, in the
form of [x, y] (or [west_east, south_north]), which indicates the end x,y location through
which the plane will pass.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
Returns An array of x,y points, which has shape num_points x 2. If xarray is enabled and the meta
parameter is True, then the result will be a xarray.DataArray object. Otherwise, the result
will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

Examples
Example 1: Using Pivot Point and Angle

from wrf import getvar, xy

from netCDF4 import Dataset

wrfnc = Dataset("wrfout_d02_2010-06-13_21:00:00")
field = wrf.getvar(wrfnc, "slp")

# Use the center of the grid

pivot = (field.shape[-1]/2.0, field.shape[-2]/2.0)

# West-East
angle = 90.0

xy_points = xy(field, pivot_point=pivot, angle=angle)

Example 2: Using Start Point and End Point

from wrf import getvar, xy

from netCDF4 import Dataset
(continues on next page)

1.6. API Reference 89

wrf-python Documentation, Release 1.2.0

(continued from previous page)

wrfnc = Dataset("wrfout_d02_2010-06-13_21:00:00")
field = wrf.getvar(wrfnc, "slp")

# Make a diagonal of lower left to upper right

start = (0, 0)
end = (-1, -1)

xy_points = xy(field, start_point=start, end_point=end)

wrf.interp1d

wrf.interp1d(field, z_in, z_out, missing=9.969209968386869e+36, meta=True)

Return the linear interpolation of a one-dimensional variable.
This function is typically used to interpolate a variable in a vertical column, but the coordinate system need not
be a vertical coordinate system. Multiple interpolation points may be specified in the z_out parameter.
Parameters
• field (xarray.DataArray or numpy.ndarray) – A one-dimensional field. Meta-
data for field is only copied to the output if field is a xarray.DataArray object.
• z_in (xarray.DataArray or numpy.ndarray) – The one-dimensional coordinates
associated with field (usually the vertical coordinates, either height or pressure).
• z_out (xarray.DataArray, numpy.ndarray) – A one-dimensional array of z_in
coordinate points to interpolate to. Must be the same type as z_in.
• missing (float, optional) – The fill value to use for the output. Default is wrf.
default_fill(np.float64).
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns An array with the same dimensionality as z_out containing the interpolated values. If xarray
is enabled and the meta parameter is True, then the result will be a xarray.DataArray
object. Otherwise, the result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

Examples
Example 1: Calculate the 850 hPa and 500 hPa values at location x,y = (100,200)

import numpy as np
from wrf import getvar, interp1d
from netCDF4 import Dataset

wrfnc = Dataset("wrfout_d02_2010-06-13_21:00:00")

(continues on next page)

90 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

(continued from previous page)

# Get a 1D vertical column for pressure at location x,y = 100,200
p_1d = wrf.getvar(wrfnc, "pres", units="hPa")[:,200,100]

# Get a 1D vertical column for height at location 100,200

ht_1d = wrf.getvar(wrfnc, "z", units="dm")[:,200,100]

# Want the heights (in decameters) at 850, 500 hPa

levels = np.asarray([850., 500.])

# Get the 850 hPa and 500 hPa values at location 100,200.
interp_vals = interp1d(p_1d, ht_1d, levels)

wrf.interp2dxy

wrf.interp2dxy(field3d, xy, meta=True)

Return a cross section for a three-dimensional field.
The returned array will hold the vertical cross section data along the line described by xy.
This method differs from wrf.vertcross() in that it will return all vertical levels found in field3d. wrf.
vertcross() includes an additional interpolation to set the output to a fixed number of vertical levels. Also,
a numpy.ma.MaskedArray is not created and this routine should be considered as low-level access to the
underlying Fortran routine.
See also:
wrf.xy(), wrf.vertcross()

Parameters
• field3d (xarray.DataArray or numpy.ndarray) – The array to interpolate with
at least three dimensions, whose rightmost dimensions are nz x ny x nx.
• xy (xarray.DataArray or numpy.ndarray) – An array of one less dimension than
field3d, whose rightmost dimensions are nxy x 2. This array holds the x,y pairs of a line
across the model domain. The requested vertical cross section will be extracted from field3d
along this line.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns An array containing the vertical cross section along the line xy. The returned dimensions
will be the same as xy, but with the rightmost dimensions being nz x nxy. If xarray is enabled and
the meta parameter is True, then the result will be a xarray.DataArray object. Otherwise,
the result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

Examples

1.6. API Reference 91

wrf-python Documentation, Release 1.2.0

Example 1: Calculate the vertical cross section for RH for a diagonal line from the lower left to the upper right
of the domain.

from wrf import getvar, xy, interp2dxy

from netCDF4 import Dataset

wrfnc = Dataset("wrfout_d02_2010-06-13_21:00:00")

rh = getvar(wrfnc, "rh")
start = (0, 0)
end = (-1, -1)
xy_line = xy(rh, start_point=start, end_point=end)

vert_cross = interp2dxy(rh, xy_line)

wrf.interpz3d

wrf.interpz3d(field3d, vert, desiredlev, missing=9.969209968386869e+36, meta=True)

Return the field interpolated to a specified pressure or height level.
This function is roughly equivalent to interplevel(), but does not handle multi-product diagnostics
(uvmet, cape_3d, etc) that contain an additional leftmost dimension for the product type. Also, a numpy.
ma.MaskedArray is not created and this routine should be considered as low-level access to the underlying
Fortran routine.
See also:
wrf.interplevel()

Parameters
• field3d (xarray.DataArray or numpy.ndarray) – A three-dimensional field to
interpolate, with the rightmost dimensions of nz x ny x nx.
• vert (xarray.DataArray or numpy.ndarray) – A three-dimensional array for the
vertical coordinate, typically pressure or height. This array must have the same dimension-
ality as field3d.
• desiredlev (float) – The desired vertical level. Must be in the same units as the vert
parameter.
• missing (float) – The fill value to use for the output. Default is wrf.
default_fill(numpy.float64).
• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns The interpolated variable. If xarray is enabled and the meta parameter is True, then the re-
sult will be an xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

92 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

Example
Example 1: Interpolate Geopotential Height to 500 hPa

from netCDF4 import Dataset

from wrf import getvar, interpz3d

wrfin = Dataset("wrfout_d02_2010-06-13_21:00:00")

p = getvar(wrfin, "pressure")
ht = getvar(wrfin, "z", units="dm")

ht_500 = interpz3d(ht, p, 500.0)

wrf.slp

wrf.slp(height, tkel, pres, qv, meta=True, units=’hPa’)

Return the sea level pressure.
This is the raw computational algorithm and does not extract any variables from WRF output files. Use wrf.
getvar() to both extract and compute diagnostic variables.
Parameters
• height (xarray.DataArray or numpy.ndarray) – Geopotential height in [m] with
the rightmost dimensions being bottom_top x south_north x west_east.
• tkel (xarray.DataArray or numpy.ndarray) – Temperature in [K] with same di-
mensionality as height.
• pres (xarray.DataArray or numpy.ndarray) – Full pressure (perturbation + base
state pressure) in [Pa] with the same dimensionality as height.

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• qv (xarray.DataArray or numpy.ndarray) – Water vapor mixing ratio in [kg/kg]

with the same dimensionality as height.
• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘slp’. Default is ‘hPa’.

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns The sea level pressure. If xarray is enabled and the meta parameter is True, then the result
will be an xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

1.6. API Reference 93

wrf-python Documentation, Release 1.2.0

See also:
wrf.getvar(), wrf.temp(), wrf.tk()

wrf.tk

wrf.tk(pres, theta, meta=True, units=’K’)

Return the temperature.
This is the raw computational algorithm and does not extract any variables from WRF output files. Use wrf.
getvar() to both extract and compute diagnostic variables.
Parameters
• pres (xarray.DataArray or numpy.ndarray) – Full pressure (perturbation + base
state pressure) in [Pa] with at least three dimensions. The rightmost dimensions are bot-
tom_top x south_north x west_east.

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• theta (xarray.DataArray or numpy.ndarray) – Potential temperature (perturba-

tion plus reference temperature) in [K] with the same dimensionality as pres.
• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘temp’. Default is ‘K’.

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns The temperature in the specified units. If xarray is enabled and the meta parameter is
True, then the result will be an xarray.DataArray object. Otherwise, the result will be a
numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.td(pres, qv, meta=True, units=’degC’)

Return the dewpoint temperature.
This is the raw computational algorithm and does not extract any variables from WRF output files. Use wrf.
getvar() to both extract and compute diagnostic variables.
Parameters
• pres (xarray.DataArray or numpy.ndarray) – Full pressure (perturbation + base
state pressure) in [hPa] with at least three dimensions. The rightmost dimensions are bot-
tom_top x south_north x west_east.

94 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• qv (xarray.DataArray or numpy.ndarray) – Water vapor mixing ratio in [kg/kg]

with the same dimensionality as pres.
• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘dp’. Default is ‘degC’.

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns The dewpoint temperature. If xarray is enabled and the meta parameter is True, then the re-
sult will be an xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.rh(qv, pres, tkel, meta=True)

Return the relative humidity.
This is the raw computational algorithm and does not extract any variables from WRF output files. Use wrf.
getvar() to both extract and compute diagnostic variables.
Parameters
• qv (xarray.DataArray or numpy.ndarray) – Water vapor mixing ratio in [kg/kg]
with at least three dimensions. The rightmost dimensions are bottom_top x south_north x
west_east.
• pres (xarray.DataArray or numpy.ndarray) – Full pressure (perturbation + base
state pressure) in [Pa] with the same dimensionality as qv.

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• tkel (xarray.DataArray or numpy.ndarray) – Temperature in [K] with same di-

mensionality as qv.

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.

1.6. API Reference 95

wrf-python Documentation, Release 1.2.0

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns The relative humidity. If xarray is enabled and the meta parameter is True, then the result
will be an xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.uvmet(u, v, lat, lon, cen_long, cone, meta=True, units=’m s-1’)

Return the u,v components of the wind rotated to earth coordinates.
The leftmost dimension of the returned array represents two different quantities:
• return_val[0,. . . ] will contain U
• return_val[1,. . . ] will contain V
This is the raw computational algorithm and does not extract any variables from WRF output files. Use wrf.
getvar() to both extract and compute diagnostic variables.
Parameters
• u (xarray.DataArray or numpy.ndarray) – The u component of the wind [m s-
1]. This variable can be staggered or unstaggered, but must be at least two dimensions. If
staggered, the rightmost dimensions are south_north x west east.
If staggered, the rightmost dimensions are south_north x west_east_stag.

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• v (xarray.DataArray or numpy.ndarray) – The v component of the wind [m s-

1]. This variable can be staggered or unstaggered, but must be at least two dimensions. If
staggered, the rightmost dimensions are south_north x west east.
If staggered, the rightmost dimensions are south_north_stag x west_east.
• lat (xarray.DataArray or numpy.ndarray) – The latitude array.
This array can either be:
– two-dimensional of size south_north x west_east.
– multi-dimensional with the same number of dimensions as u and v, but with rightmost
dimensions south_north x west_east and the same leftmost dimensions as u and v
– multi-dimensional with one fewer dimensions as u and v, with rightmost dimensions
south_north x west_east and the same leftmost dimensions as u and v, minus the third-
from-the-right dimension of u and v.

96 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

Note: This variable must also be supplied as a xarray.DataArray in order to copy the
dimension names to the output. Otherwise, default names will be used.

• lon (xarray.DataArray or numpy.ndarray) – The longitude array.

This array can either be:
– two-dimensional of size south_north x west_east.
– multi-dimensional with the same number of dimensions as u and v, but with rightmost
dimensions south_north x west_east and the same leftmost dimensions as u and v
– multi-dimensional with one fewer dimensions as u and v, with rightmost dimensions
south_north x west_east and the same leftmost dimensions as u and v, minus the third-
from-the-right dimension of u and v.
• cen_long (float) – The standard longitude for the map projection.
• cone (float) – The cone factor used for the map project. If the projection is not a conic
projection, the cone is simply 1.0. For conic projections, the cone factor is given by:

if((fabs(true_lat1 - true_lat2) > 0.1) and

(fabs(true_lat2 - 90.) > 0.1)):
cone = (log(cos(true_lat1*radians_per_degree))
- log(cos(true_lat2*radians_per_degree)))

cone = (cone /
(log(tan((45.-fabs(true_lat1/2.))*radians_per_degree))
- log(tan((45.-fabs(true_lat2/2.))*radians_per_degree))))
else:
cone = sin(fabs(true_lat1)*radians_per_degree)

• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘uvmet’. Default is ‘m s-1’.

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns The u,v components of the wind rotated to earth coordinates. The leftmost dimension size
is 2, for u and v. If xarray is enabled and the meta parameter is True, then the result will be an
xarray.DataArray object. Otherwise, the result will be a numpy.ndarray object with
no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.smooth2d(field, passes, meta=True)

Return the field smoothed.

1.6. API Reference 97

wrf-python Documentation, Release 1.2.0

This routine does not modify the original data.

Parameters
• field (xarray.DataArray or numpy.ndarray) – The field to smooth, which must
be at least two dimensions. Missing/fill values will be ignored as long as the type is either
a numpy.ma.MaskedArray or a :class:`xarray.DataArray with a _Fill-
Value attribute.
• passes (int) – The number of smoothing passes.
• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.
Returns The smoothed field. If xarray is enabled and the meta parameter is True, then the result will
be an xarray.DataArray object. Otherwise, the result will be either a numpy.ndarray
or a numpy.ma.MaskedArray depending on the type for field.
Return type xarray.DataArray, numpy.ma.MaskedArray or numpy.ndarray)

wrf.cape_2d

wrf.cape_2d(pres_hpa, tkel, qv, height, terrain, psfc_hpa, ter_follow, missing=9.969209968386869e+36,

meta=True)
Return the two-dimensional MCAPE, MCIN, LCL, and LFC.
This function calculates the maximum convective available potential energy (MCAPE), maximum convective
inhibition (MCIN), lifted condensation level (LCL), and level of free convection (LFC). This function uses the
RIP [Read/Interpolate/plot] code to calculate potential energy (CAPE) and convective inhibition (CIN) [J kg-1]
only for the parcel with max theta-e in the column (i.e. something akin to Colman’s MCAPE). CAPE is defined
as the accumulated buoyant energy from the level of free convection (LFC) to the equilibrium level (EL). CIN
is defined as the accumulated negative buoyant energy from the parcel starting point to the LFC.
The cape_2d algorithm works by first finding the maximum theta-e height level in the lowest 3000 m. A parcel
with a depth of 500 m is then calculated and centered over this maximum theta-e height level. The parcel’s
moisture and temperature characteristics are calculated by averaging over the depth of this 500 m parcel. This
‘maximum’ parcel is then used to compute MCAPE, MCIN, LCL and LFC.
The leftmost dimension of the returned array represents four different quantities:
• return_val[0,. . . ] will contain MCAPE [J kg-1]
• return_val[1,. . . ] will contain MCIN [J kg-1]
• return_val[2,. . . ] will contain LCL [m]
• return_val[3,. . . ] will contain LFC [m]
This function also supports computing MCAPE along a single vertical column. In this mode, the pres_hpa, tkel,
qv and height arguments must be one-dimensional vertical columns, and the terrain and psfc_hpa arguments
must be scalar values (float, numpy.float32 or numpy.float64).
This is the raw computational algorithm and does not extract any variables from WRF output files. Use wrf.
getvar() to both extract and compute diagnostic variables.
Parameters
• pres_hpa (xarray.DataArray or numpy.ndarray) – Full pressure (perturbation
+ base state pressure) in [hPa] with at least three dimensions. The rightmost dimensions
can be top_bottom x south_north x west_east or bottom_top x south_north x west_east.
When operating on only a single column of values, the vertical column can be bottom_top
or top_bottom. In this case, terrain and psfc_hpa must be scalars.

98 Chapter 1. Documentation
wrf-python Documentation, Release 1.2.0

Note: The units for pres_hpa are [hPa].

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• tkel (xarray.DataArray or numpy.ndarray) – Temperature in [K] with same di-

mensionality as pres_hpa.
• qv (xarray.DataArray or numpy.ndarray) – Water vapor mixing ratio in [kg/kg]
with the same dimensionality as pres_hpa.
• height (xarray.DataArray or numpy.ndarray) – Geopotential height in [m] with
the same dimensionality as pres_hpa.
• terrain (xarray.DataArray or numpy.ndarray) – Terrain height in [m]. This is
at least a two-dimensional array with the same dimensionality as pres_hpa, excluding the
vertical (bottom_top/top_bottom) dimension. When operating on a single vertical column,
this argument must be a scalar (float, numpy.float32, or numpy.float64).
• psfc_hpa (xarray.DataArray or numpy.ndarray) – The surface pressure in
[hPa]. This is at least a two-dimensional array with the same dimensionality as pres_hpa,
excluding the vertical (bottom_top/top_bottom) dimension. When operating on a singlev-
ertical column, this argument must be a scalar (float, numpy.float32, or numpy.
float64).

Note: The units for psfc_hpa are [hPa].

• ter_follow (bool) – A boolean that should be set to True if the data uses terrain fol-
lowing coordinates (WRF data). Set to False for pressure level data.
• missing (float, optional) – The fill value to use for the output. Default is wrf.
default_fill(numpy.float64).
• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns The cape, cin, lcl, and lfc values as an array whose leftmost dimension is 4 (0=CAPE,
1=CIN, 2=LCL, 3=LFC) . If xarray is enabled and the meta parameter is True, then the result
will be an xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

1.6. API Reference 99

wrf-python Documentation, Release 1.2.0

wrf.cape_3d

wrf.cape_3d(pres_hpa, tkel, qv, height, terrain, psfc_hpa, ter_follow, missing=9.969209968386869e+36,

meta=True)
Return the three-dimensional CAPE and CIN.
This function calculates the maximum convective available potential energy (CAPE) and maximum convective
inhibition (CIN). This function uses the RIP [Read/Interpolate/plot] code to calculate potential energy (CAPE)
and convective inhibition (CIN) [J kg-1] for every grid point in the entire 3D domain (treating each grid point
as a parcel).
The leftmost dimension of the returned array represents two different quantities:
• return_val[0,. . . ] will contain CAPE [J kg-1]
• return_val[1,. . . ] will contain CIN [J kg-1]
This function also supports computing CAPE along a single vertical column. In this mode, the pres_hpa, tkel,
qv and height arguments must be one-dimensional vertical columns, and the terrain and psfc_hpa arguments
must be scalar values (float, numpy.float32 or numpy.float64).
This is the raw computational algorithm and does not extract any variables from WRF output files. Use wrf.
getvar() to both extract and compute diagnostic variables.
Parameters
• pres_hpa (xarray.DataArray or numpy.ndarray) – Full pressure (perturbation
+ base state pressure) in [hPa] with at least three dimensions when operating on a grid of
values. The rightmost dimensions can be top_bottom x south_north x west_east or bot-
tom_top x south_north x west_east. When operating on only a single column of values, the
vertical column can be bottom_top or top_bottom. In this case, terrain and psfc_hpa must
be scalars.

Note: The units for pres_hpa are [hPa].

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• tkel (xarray.DataArray or numpy.ndarray) – Temperature in [K] with same di-

mensionality as pres_hpa.
• qv (xarray.DataArray or numpy.ndarray) – Water vapor mixing ratio in [kg/kg]
with the same dimensionality as pres_hpa.
• height (xarray.DataArray or numpy.ndarray) – Geopotential height in [m] with
the same dimensionality as pres_hpa.
• terrain (xarray.DataArray, numpy.ndarray, or a scalar) – Terrain height in
[m]. When operating on a grid of values, this argument is at least a two-dimensional array
with the same dimensionality as pres_hpa, excluding the vertical (bottom_top/top_bottom)
dimension. When operating on a single vertical column, this argument must be a scalar
(float, numpy.float32, or numpy.float64).
• psfc_hpa (xarray.DataArray, numpy.ndarray, or a scalar) – Surface pressure in
[hPa]. When operating on a grid of values, this argument is at least a two-dimensional array
with the same dimensionality as pres_hpa, excluding the vertical (bottom_top/top_bottom)

100 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

dimension. When operating on a single vertical column, this argument must be a scalar
(float, numpy.float32, or numpy.float64).

Note: The units for psfc_hpa are [hPa].

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns The CAPE and CIN as an array whose leftmost dimension is 2 (0=CAPE, 1=CIN). If xarray
is enabled and the meta parameter is True, then the result will be an xarray.DataArray
object. Otherwise, the result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.cloudfrac(vert, relh, vert_inc_w_height, low_thresh, mid_thresh, high_thresh, miss-

ing=9.969209968386869e+36, meta=True)
Return the cloud fraction.
The leftmost dimension of the returned array represents three different quantities:
• return_val[0,. . . ] will contain LOW level cloud fraction
• return_val[1,. . . ] will contain MID level cloud fraction
• return_val[2,. . . ] will contain HIGH level cloud fraction
The low_thresh, mid_thresh, and high_threshold paramters specify the low, mid, and high cloud levels in the
same units as vert.
In mountainous regions, there is a possibility that the lowest WRF level will be higher than the low_cloud or
mid_cloud threshold. When this happens, a fill value will be used in the output at that location.
This is the raw computational algorithm and does not extract any variables from WRF output files. Use wrf.
getvar() to both extract and compute diagnostic variables.
Parameters
• vert (xarray.DataArray or numpy.ndarray) – The vertical coordinate variable
(usually pressure or height) with the rightmost dimensions as bottom_top x south_north x
west_east

1.6. API Reference 101

wrf-python Documentation, Release 1.2.0

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• relh (xarray.DataArray or numpy.ndarray) – Relative humidity with the same

dimensionality as vert
• vert_inc_w_height (int) – Set to 1 if the vertical coordinate values increase with
height (height values). Set to 0 if the vertical coordinate values decrease with height (pres-
sure values).
• low_thresh (float) – The bottom vertical threshold for what is considered a low cloud.
• mid_thresh (float) – The bottom vertical threshold for what is considered a mid level
cloud.
• high_thresh (float) – The bottom vertical threshold for what is considered a high
cloud.
• missing (float:, optional) – The fill value to use for areas where the sur-
face is higher than the cloud threshold level (e.g. mountains). Default is wrf.
default_fill(numpy.float64).
• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns The cloud fraction array whose leftmost dimension is 3 (LOW=0, MID=1, HIGH=2).
If xarray is enabled and the meta parameter is True, then the result will be an xarray.
DataArray object. Otherwise, the result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.ctt(pres_hpa, tkel, qv, qcld, height, terrain, qice=None, fill_nocloud=False, miss-

ing=9.969209968386869e+36, opt_thresh=1.0, meta=True, units=’degC’)
Return the cloud top temperature.
This is the raw computational algorithm and does not extract any variables from WRF output files. Use wrf.
getvar() to both extract and compute diagnostic variables.
Parameters
• pres_hpa (xarray.DataArray or numpy.ndarray) – Full pressure (perturbation
+ base state pressure) in [hPa], with the rightmost dimensions as bottom_top x south_north
x west_east

Note: The units for psfc_hpa are [hPa].

102 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• tkel (xarray.DataArray or numpy.ndarray) – Temperature in [K] with same di-

mensionality as pres_hpa.
• qv (xarray.DataArray or numpy.ndarray) – Water vapor mixing ratio in [kg/kg]
with the same dimensionality as pres_hpa.
• qcld (xarray.DataArray or numpy.ndarray) – Cloud water vapor mixing ratio in
[kg/kg] with the same dimensionality as pres_hpa.
• height (xarray.DataArray or numpy.ndarray) – Geopotential height in [m] with
the same dimensionality as pres_hpa.
• terrain (xarray.DataArray or numpy.ndarray) – Terrain height in [m]. This is
at least a two-dimensional array with the same dimensionality as pres_hpa, excluding the
vertical (bottom_top/top_bottom) dimension.
• qice (xarray.DataArray or numpy.ndarray, optional) – Ice mixing ratio in
[kg/kg] with the same dimensionality as pres_hpa.
• fill_nocloud (bool, optional) – Set to True to use fill values in regions where clouds
are not detected (optical depth less than 1). Otherwise, the output will contain the surface
temperature for areas without clouds. Default is False.
• missing (float, optional) – The fill value to use for areas where no clouds are detected.
Only used if fill_nocloud is True. Default is wrf.default_fill(numpy.float64).
• opt_thresh (float, optional) – The amount of optical depth (integrated from top down)
required to trigger a cloud top temperature calculation. The cloud top temperature is cal-
culated at the vertical level where this threshold is met. Vertical columns with less than
this threshold will be treated as cloud free areas. In general, the larger the value is for this
threshold, the lower the altitude will be for the cloud top temperature calculation, and there-
fore higher cloud top temperature values. Default is 1.0, which should be sufficient for most
users.
• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘ctt’. Default is ‘degC’.

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns The cloud top temperature. If xarray is enabled and the meta parameter is True, then
the result will be an xarray.DataArray object. Otherwise, the result will be a numpy.
ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

1.6. API Reference 103

wrf-python Documentation, Release 1.2.0

wrf.dbz

wrf.dbz(pres, tkel, qv, qr, qs=None, qg=None, use_varint=False, use_liqskin=False, meta=True)

Return the simulated radar reflectivity.
This function computes equivalent reflectivity factor [dBZ] at each model grid point assuming spherical particles
of constant density, with exponential size distributions. This function is based on “dbzcalc.f” in RIP.
This is the raw computational algorithm and does not extract any variables from WRF output files. Use wrf.
getvar() to both extract and compute diagnostic variables.
Parameters
• pres (xarray.DataArray or numpy.ndarray) – Full pressure (perturbation + base
state pressure) in [Pa], with the rightmost dimensions as bottom_top x south_north x
west_east

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• tkel (xarray.DataArray or numpy.ndarray) – Temperature in [K] with same di-

mensionality as pres.
• qv (xarray.DataArray or numpy.ndarray) – Water vapor mixing ratio in [kg/kg]
with the same dimensionality as pres.
• qr (xarray.DataArray or numpy.ndarray) – Rain water vapor mixing ratio in
[kg/kg] with the same dimensionality as pres.
• qs (xarray.DataArray or numpy.ndarray, optional) – Snow mixing ratio in
[kg/kg] with the same dimensionality as pres.
• qg (xarray.DataArray or numpy.ndarray, optional) – Graupel mixing ratio in
[kg/kg] with the same dimensionality as pres.
• use_varint (bool, optional) – When set to False, the intercept parameters are assumed
constant (as in MM5’s Reisner-2 bulk microphysical scheme). When set to True, the vari-
able intercept parameters are used as in the more recent version of Reisner-2 (based on
Thompson, Rasmussen, and Manning, 2004, Monthly weather Review, Vol. 132, No. 2, pp.
519-542.).
• use_liqskin (bool, optional) – When set to True, frozen particles that are at a temper-
ature above freezing are assumed to scatter as a liquid particle. Set to False to disable.
• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns The simulated radar reflectivity. If xarray is enabled and the meta parameter is True, then
the result will be an xarray.DataArray object. Otherwise, the result will be a numpy.
ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

104 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

wrf.srhel

wrf.srhel(u, v, height, terrain, top=3000.0, meta=True)

Return the storm relative helicity.
This function calculates storm relative helicity from WRF ARW output. SRH (Storm Relative Helicity) is
a measure of the potential for cyclonic updraft rotation in right-moving supercells, and is calculated for the
lowest 1-km and 3-km layers above ground level. There is no clear threshold value for SRH when forecasting
supercells, since the formation of supercells appears to be related more strongly to the deeper layer vertical shear.
Larger values of 0-3 km SRH (greater than 250 m2 s-2) and 0-1 km SRH (greater than 100 m2 s-2), however, do
suggest an increased threat of tornadoes with supercells. For SRH, larger values are generally better, but there
are no clear “boundaries” between non-tornadic and significant tornadic supercells.
This is the raw computational algorithm and does not extract any variables from WRF output files. Use wrf.
getvar() to both extract and compute diagnostic variables.
Parameters
• u (xarray.DataArray or numpy.ndarray) – The u component of the wind that must
have at least three dimensions. The rightmost dimensions are bottom_top x south_north x
west_east.
• v (xarray.DataArray or numpy.ndarray) – The v component of the wind with the
same dimensionality as u.
• height (xarray.DataArray or numpy.ndarray) – Geopotential height in [m] with
the same dimensionality as u.
• terrain (xarray.DataArray or numpy.ndarray) – Terrain height in [m]. This
is at least a two-dimensional array with the same dimensionality as u, excluding the bot-
tom_top dimension.

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• top (float) – The height of the layer below which helicity is calculated (meters above
ground level).
• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns The storm relative helicity. If xarray is enabled and the meta parameter is True, then the re-
sult will be an xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

1.6. API Reference 105

wrf-python Documentation, Release 1.2.0

wrf.udhel

wrf.udhel(zstag, mapfct, u, v, wstag, dx, dy, bottom=2000.0, top=5000.0, meta=True)

Return the updraft helicity.
This function calculates updraft helicity to detect rotating updrafts. The formula follows Kain et al., 2008, Wea.
and Forecasting, 931-952, but this version has controls for the limits of integration, bottom to top, in m AGL.
Kain et al used 2000 to 5000 m. The expected range is 25 to 250 m-2/s-2. Keith Brewster, CAPS/Univ. of
Oklahoma ; March, 2010
This is the raw computational algorithm and does not extract any variables from WRF output files. Use wrf.
getvar() to both extract and compute diagnostic variables.
Parameters
• zstag (xarray.DataArray or numpy.ndarray) – Geopotential height in [m] that
is at least three dimensions with a staggered vertical dimension. The rightmost dimensions
are bottom_top_stag x south_north x west_east.
• mapfct (xarray.DataArray or numpy.ndarray) – The map scale factor on the
mass grid. An array of at least two dimensions, whose rightmost two dimensions must be
south_north x west_east. If this array is more than two dimensions, they must be the same
as zstag’s leftmost dimensions.
• u (xarray.DataArray or numpy.ndarray) – The u component of the wind [m s-
1] whose rightmost three dimensions must be bottom_top x south_north x west_east. The
leftmost dimensions must be the same as zp’s leftmost dimensions.

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• v (xarray.DataArray or numpy.ndarray) – The v component of the wind [m s-

1] whose rightmost three dimensions must be bottom_top x south_north x west_east. The
leftmost dimensions must be the same as zstag’s leftmost dimensions.
• wstag (xarray.DataArray or numpy.ndarray) – The z component of the wind [m
s-1] with the same dimensionality as zstag.
• dx (float) – The distance between x grid points.
• dy (float) – The distance between y grid points.
• bottom (float, optional) – The bottom limit of integration. Default is 2000.0.
• top (float, optional) – The upper limit of integration. Default is 5000.0.
• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns The updraft helicity. If xarray is enabled and the meta parameter is True, then the result
will be an xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

106 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

wrf.avo(ustag, vstag, msfu, msfv, msfm, cor, dx, dy, meta=True)

Return the absolute vorticity.
This function returns absolute vorticity [10-5 s-1], which is the sum of the relative vorticity at each grid point
and the Coriolis parameter at the latitude.
This is the raw computational algorithm and does not extract any variables from WRF output files. Use wrf.
getvar() to both extract and compute diagnostic variables.
Parameters
• ustag (xarray.DataArray or numpy.ndarray) – The u component of the wind in
[m s-1] that is at least three dimensions with a staggered west_east dimension. The rightmost
dimensions are bottom_top x south_north x west_east_stag.

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• vstag (xarray.DataArray or numpy.ndarray) – The v component of the wind

in [m s-1] that is at least three dimensions with a staggered south_north dimension. The
rightmost dimensions are bottom_top x south_north_stag x west_east.

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• msfu (xarray.DataArray or numpy.ndarray) – The map scale factor on the u-grid

that is at least two dimensions, whose rightmost two dimensions must be the same as ustag.
If this array contains more than two dimensions, they must be the same as ustag and vstag’s
leftmost dimensions.
• msfv (xarray.DataArray or numpy.ndarray) – The map scale factor on the v-grid
that is at least two dimensions, whose rightmost two dimensions must be the same as vstag.
If this array contains more than two dimensions, they must be the same as ustag and vstag’s
leftmost dimensions.
• msfm (xarray.DataArray or numpy.ndarray) – The map scale factor on the mass
grid that is at least two dimensions, whose rightmost two dimensions must be south_north x
west_east. If this array contains more than two dimensions, they must be the same as ustag
and vstag’s leftmost dimensions.
• cor (xarray.DataArray or numpy.ndarray) – The Coriolis sine latitude array that
is at least two dimensions, whose dimensions must be the same as msfm.
• dx (float) – The distance between x grid points.
• dy (float) – The distance between y grid points.
• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.

1.6. API Reference 107

wrf-python Documentation, Release 1.2.0

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns The absolute vorticity. If xarray is enabled and the meta parameter is True, then the result
will be an xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• pres (xarray.DataArray or numpy.ndarray) – Full pressure (perturbation + base

state pressure) in [Pa], with the same dimensions as theta.
• msfu (xarray.DataArray or numpy.ndarray) – The map scale factor on the u-grid
that is at least two dimensions, whose rightmost two dimensions must be the same as ustag.
If this array contains more than two dimensions, they must be the same as ustag and vstag’s
leftmost dimensions.
• msfv (xarray.DataArray or numpy.ndarray) – The map scale factor on the v-grid
that is at least two dimensions, whose rightmost two dimensions must be the same as vstag.
If this array contains more than two dimensions, they must be the same as ustag and vstag’s
leftmost dimensions.
• msfm (xarray.DataArray or numpy.ndarray) – The map scale factor on the mass
grid that is at least two dimensions, whose rightmost two dimensions must be south_north x
west_east. If this array contains more than two dimensions, they must be the same as ustag
and vstag’s leftmost dimensions.

108 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

• cor (xarray.DataArray or numpy.ndarray) – The Coriolis sine latitude array that

is at least two dimensions, whose dimensions must be the same as msfm.
• dx (float) – The distance between x grid points.
• dy (float) – The distance between y grid points.
• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns The potential vorticity. If xarray is enabled and the meta parameter is True, then the result
will be an xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.eth(qv, tkel, pres, meta=True, units=’K’)

Return the equivalent potential temperature.
This is the raw computational algorithm and does not extract any variables from WRF output files. Use wrf.
getvar() to both extract and compute diagnostic variables.
Parameters
• qv (xarray.DataArray or numpy.ndarray) – Water vapor mixing ratio in [kg/kg]
that is at least three dimensions, with the rightmost dimensions of bottom_top x south_north
x west_east.
• tkel (xarray.DataArray or numpy.ndarray) – Temperature in [K] with same di-
mensionality as qv.

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• pres (xarray.DataArray or numpy.ndarray) – Full pressure (perturbation + base

state pressure) in [Pa] with the same dimensionality as qv.
• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘eth’. Default is ‘K’.

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

1.6. API Reference 109

wrf-python Documentation, Release 1.2.0

Returns The equivalent potential temperature. If xarray is enabled and the meta parameter is True,
then the result will be an xarray.DataArray object. Otherwise, the result will be a numpy.
ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

See also:
wrf.getvar(), wrf.temp(), wrf.wetbulb(), tvirtual()

wrf.wetbulb

wrf.wetbulb(pres, tkel, qv, meta=True, units=’K’)

Return the wetbulb temperature.
This is the raw computational algorithm and does not extract any variables from WRF output files. Use wrf.
getvar() to both extract and compute diagnostic variables.
Parameters
• pres (xarray.DataArray or numpy.ndarray) – Full pressure (perturbation + base
state pressure) in [Pa], with the rightmost dimensions as bottom_top x south_north x
west_east

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• tkel (xarray.DataArray or numpy.ndarray) – Temperature in [K] with same di-

mensionality as pres.
• qv (xarray.DataArray or numpy.ndarray) – Water vapor mixing ratio in [kg/kg]
with the same dimensionality as pres
• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘twb’. Default is ‘K’.

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns The wetbulb temperature. If xarray is enabled and the meta parameter is True, then the re-
sult will be an xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

See also:
wrf.getvar(), wrf.temp(), wrf.eth(), tvirtual()

wrf.tvirtual

wrf.tvirtual(tkel, qv, meta=True, units=’K’)

Return the virtual temperature.

110 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

This is the raw computational algorithm and does not extract any variables from WRF output files. Use wrf.
getvar() to both extract and compute diagnostic variables.
Parameters
• tkel (xarray.DataArray or numpy.ndarray) – Temperature in [K] with the right-
most dimensions as bottom_top x south_north x west_east.

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• qv (xarray.DataArray or numpy.ndarray) – Water vapor mixing ratio in [kg/kg]

with the same dimensionality as tkel
• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘tv’. Default is ‘K’.

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns The virtual temperature. If xarray is enabled and the meta parameter is True, then the result
will be an xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

See also:
wrf.getvar(), wrf.temp(), wrf.eth(), wetbulb()

wrf.omega

wrf.omega(qv, tkel, w, pres, meta=True)

Return omega.
This function calculates omega (dp/dt) [Pa s-1].
This is the raw computational algorithm and does not extract any variables from WRF output files. Use wrf.
getvar() to both extract and compute diagnostic variables.
Parameters
• qv (xarray.DataArray or numpy.ndarray) – Water vapor mixing ratio in [kg/kg]
with the rightmost dimensions as bottom_top x south_north x west_east.

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• tkel (xarray.DataArray or numpy.ndarray) – Temperature in [K] with the same

dimensionality as qv.
• w (xarray.DataArray or numpy.ndarray) – The vertical velocity [m s-1] with the
same dimensionality as qv.

1.6. API Reference 111

wrf-python Documentation, Release 1.2.0

• pres (xarray.DataArray or numpy.ndarray) – Full pressure (perturbation + base

state pressure) in [Pa] with the same dimensionality as qv.
• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns Omega. If xarray is enabled and the meta parameter is True, then the result will be an
xarray.DataArray object. Otherwise, the result will be a numpy.ndarray object with
no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.pw(pres, tkel, qv, height, meta=True)

Return the precipitable water.
This is the raw computational algorithm and does not extract any variables from WRF output files. Use wrf.
getvar() to both extract and compute diagnostic variables.
Parameters
• pres (xarray.DataArray or numpy.ndarray) – Full pressure (perturbation + base
state pressure) in [Pa], with the rightmost dimensions as bottom_top x south_north x
west_east

Note: This variable must be supplied as a xarray.DataArray in order to copy the

dimension names to the output. Otherwise, default names will be used.

• tkel (xarray.DataArray or numpy.ndarray) – Temperature in [K] with the same

dimensionality as pres.
• qv (xarray.DataArray or numpy.ndarray) – Water vapor mixing ratio in [kg/kg]
with the same dimensionality as pres
• height (xarray.DataArray or numpy.ndarray) – Geopotential height in [m] with
the same dimensionality as pres.
• meta (bool) – Set to False to disable metadata and return numpy.ndarray instead of
xarray.DataArray. Default is True.

Warning: The input arrays must not contain any missing/fill values or numpy.nan values.

Returns The precipitable water [kg m-2]. If xarray is enabled and the meta parameter is True, then
the result will be an xarray.DataArray object. Otherwise, the result will be a numpy.
ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

112 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

OpenMP Runtime Library Routines

The routines below are the OpenMP runtime libraries that have been wrapped for wrf-python. The entire library
(OpenMP 3.x) has been wrapped, but many of the routines are only useful inside of an OpenMP thread, so they aren’t
useful from inside the Python interpreter. Also, the Fortran code in wrf-python is fairly simple in terms of threading,
so features like nested threads aren’t used. The documentation below is split in to the useful OpenMP functions and
the less useful functions.
The documentation for each routine was taken directly from the OpenMP Specification. Read the specification for
more details about these routines.

Useful OpenMP Routines

The routines below are useful when called from within a Python program. These routines handle setting the number
of threads, setting up the scheduler, and timing.
It is also important to note that the OpenMP directives within the Fortran code all specify a runtime scheduler. This
means that the user can control the type of scheduling to use from within their Python application by using the routines
below.

wrf.omp_enabled Return True if OpenMP is enabled.

wrf.omp_set_num_threads Specify the number of threads to use.
wrf.omp_get_max_threads Return the maximum number of threads that can be used
in a parallel region.
wrf.omp_get_num_procs Return the number of processors on the device.
wrf.omp_set_dynamic Enable or disable dynamic adjustment of the number of
threads available for the execution of subsequent paral-
lel regions by setting the value of the dyn-var ICV.
wrf.omp_get_dynamic Return the value of the dyn-var ICV, which determines
whether dynamic adjustment of the number of threads
is enabled or disabled.
wrf.omp_set_schedule Set the schedule that is applied when runtime is used as
schedule kind, by setting the value of the run-sched-var
ICV.
wrf.omp_get_schedule Return the schedule that is applied when the runtime
schedule is used.
wrf.omp_get_thread_limit Return the maximum number of OpenMP threads avail-
able to participate in the current contention group.
wrf.omp_get_wtime Return elapsed wall clock time in seconds.
wrf.omp_get_wtick Return the precision of the timer used by wrf.
omp_get_wtime().

wrf.omp_enabled

wrf.omp_enabled()
Return True if OpenMP is enabled.
OpenMP is only enabled if compiled with OpenMP features.

1.6. API Reference 113

wrf-python Documentation, Release 1.2.0

Returns True if OpenMP is enabled, otherwise False.

Return type bool

wrf.omp_set_num_threads

wrf.omp_set_num_threads(num_threads)
Specify the number of threads to use.
The omp_set_num_threads routine affects the number of threads to be used for subsequent parallel regions that
do not specify a num_threads clause, by setting the value of the first element of the nthreads-var ICV of the
current task.
Parameters num_threads (a positive int) – The number of threads. Must be positive.
Returns None.

wrf.omp_get_max_threads

wrf.omp_get_max_threads()
Return the maximum number of threads that can be used in a parallel region.
The omp_get_max_threads routine returns an upper bound on the number of threads that could be used to form
a new team if a parallel construct without a num_threads clause were encountered after execution returns from
this routine.
Returns The number of threads in the current team.
Return type int
See also:
wrf.omp_set_num_threads()

wrf.omp_get_num_procs

wrf.omp_get_num_procs()
Return the number of processors on the device.
The omp_get_num_procs routine returns the number of processors that are available to the device at the time
the routine is called. This value may change between the time that it is determined by the omp_get_num_procs
routine and the time that it is read in the calling context due to system actions outside the control of the OpenMP
implementation.
Returns The number of processors.
Return type int

wrf.omp_set_dynamic

wrf.omp_set_dynamic(dynamic_threads)
Enable or disable dynamic adjustment of the number of threads available for the execution of subsequent parallel
regions by setting the value of the dyn-var ICV.
For implementations that support dynamic adjustment of the number of threads, if the argument to
omp_set_dynamic evaluates to True, dynamic adjustment is enabled for the current task; otherwise, dynamic

114 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

adjustment is disabled for the current task. For implementations that do not support dynamic adjustment of the
number of threads this routine has no effect: the value of dyn-var remains false.
Parameters dynamic_threads (bool) – Set to True to support the dynamic adjustment of the
number of threads. Otherwise, set to False.
Returns None.
See also:
wrf.omp_get_dynamic()

wrf.omp_get_dynamic

wrf.omp_get_dynamic()
Return the value of the dyn-var ICV, which determines whether dynamic adjustment of the number of threads is
enabled or disabled.
This routine returns 1 if dynamic adjustment of the number of threads is enabled for the current task; it returns
0, otherwise. If an implementation does not support dynamic adjustment of the number of threads, then this
routine always returns 0.
Returns Returns 1 if dynamic thread adjustment is enabled, 0 if disabled.
Return type int
See also:
wrf.omp_set_dynamic()

wrf.omp_set_schedule

wrf.omp_set_schedule(kind, modifier=0)
Set the schedule that is applied when runtime is used as schedule kind, by setting the value of the run-sched-var
ICV.
The effect of this routine is to set the value of the run-sched-var ICV of the current task to the values specified
in the two arguments. The schedule is set to the schedule type specified by the first argument kind. It can
be any of the standard schedule types or any other implementation specific one. For the schedule types static,
dynamic, and guided the chunk_size is set to the value of the second argument, or to the default chunk_size if
the value of the second argument is less than 1; for the schedule type auto the second argument has no meaning;
for implementation specific schedule types, the values and associated meanings of the second argument are
implementation defined.
Parameters
• kind (int) – Must be wrf.OMP_SCHED_STATIC, wrf.OMP_SCHED_DYNAMIC,
wrf.OMP_SCHED_GUIDED, or wrf.OMP_SCHED_AUTO.
• modifier (int) – An implementation specific value, depending on the choice for kind.
This parameter is alternatively named chunk_size in some OpenMP documentation. Default
is 0, which means the OpenMP implementation will use its default value.
Returns None
See also:
wrf.omp_get_schedule()

1.6. API Reference 115

wrf-python Documentation, Release 1.2.0

wrf.omp_get_schedule

wrf.omp_get_schedule()
Return the schedule that is applied when the runtime schedule is used.
This routine returns the run-sched-var ICV in the task to which the routine binds. The first item is the
schedule kind, which will be one of wrf.OMP_SCHED_STATIC, wrf.OMP_SCHED_DYNAMIC, wrf.
OMP_SCHED_GUIDED, or wrf.OMP_SCHED_AUTO. The second item returned is the modifier, which is often
named chunk_size in OpenMP documentation.
Returns The first item is an int for the schedule kind. The second items is int for the modifier
(chunk_size).
Return type tuple
See also:
wrf.omp_set_schedule()

wrf.omp_get_thread_limit

wrf.omp_get_thread_limit()
Return the maximum number of OpenMP threads available to participate in the current contention group.
The omp_get_thread_limit routine returns the value of the thread-limit-var ICV.
Returns The number of OpenMP threads available to participate in the current contention group.
Return type int
See also:
wrf.omp_get_max_threads()

wrf.omp_get_wtime

wrf.omp_get_wtime()
Return elapsed wall clock time in seconds.
The omp_get_wtime routine returns a value equal to the elapsed wall clock time in seconds since some “time
in the past”. The actual “time in the past” is arbitrary, but it is guaranteed not to change during the execution of
the application program. The time returned is a “per-thread time”, so it is not required to be globally consistent
across all threads participating in an application.
Returns Returns the wall clock time in seconds.
Return type float
See also:
wrf.omp_get_wtick()

wrf.omp_get_wtick

wrf.omp_get_wtick()
Return the precision of the timer used by wrf.omp_get_wtime().
The omp_get_wtick routine returns a value equal to the number of seconds between successive clock ticks of
the timer used by wrf.omp_get_wtime().

116 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

Returns Returns the precision of the timer.

Return type float
See also:
wrf.omp_get_wtime()

Less Useful OpenMP Routines

The routines below are less useful because wrf-python does not use nested parallelism and some of the routines are
only applicable when called from within an OpenMP thread.

wrf.omp_get_num_threads Return the number of threads in the current team.

wrf.omp_get_thread_num Return the thread number, within the current team, of
the calling thread.
wrf.omp_in_parallel Return 1 if the active-levels-var ICV is greater than zero;
otherwise, return 0.
wrf.omp_set_nested Enable or disable nested parallelism, by setting the nest-
var ICV
wrf.omp_get_nested Return the value of the nest-var ICV, which determines
if nested parallelism is enabled or disabled
wrf.omp_set_max_active_levels Limit the number of nested active parallel regions on the
device, by setting the max-active-levels-var ICV.
wrf.omp_get_max_active_levels Return the value of the max-active-levels-var ICV,
which determines the maximum number of nested ac-
tive parallel regions on the device
wrf.omp_get_level Return the value of the levels-var ICV.
wrf.omp_get_ancestor_thread_num Return, for a given nested level of the current thread, the
thread number of the ancestor of the current thread.
wrf.omp_get_team_size Return, for a given nested level of the current thread,
the size of the thread team to which the ancestor or the
current thread belongs
wrf.omp_get_active_level Return the value of the active-level-vars ICV.
wrf.omp_in_final Return 1 (True) if the routine is executed in a final task
region; otherwise, it returns 0 (False).
wrf.omp_init_lock Initialize a simple OpenMP lock.
wrf.omp_init_nest_lock Initialize a nestable OpenMP lock.
wrf.omp_destroy_lock Destroy a simple OpenMP lock.
wrf.omp_destroy_nest_lock Destroy a nestable OpenMP lock.
wrf.omp_set_lock Set a simple OpenMP lock.
wrf.omp_set_nest_lock Set a nestable OpenMP lock.
wrf.omp_unset_lock Unset a simple OpenMP lock.
wrf.omp_unset_nest_lock Unset a nestable OpenMP lock.
wrf.omp_test_lock Test a simple OpenMP lock.
wrf.omp_test_nest_lock Test a nestable OpenMP lock.

wrf.omp_get_num_threads

wrf.omp_get_num_threads()
Return the number of threads in the current team.

1.6. API Reference 117

wrf-python Documentation, Release 1.2.0

The omp_get_num_threads routine returns the number of threads in the team executing the parallel region to
which the routine region binds. If called from the sequential part of a program, this routine returns 1.

Note: This function always returns 1 when called from within Python.

Returns The number of threads in the current team.

Return type int

See also:
wrf.omp_get_max_threads(), wrf.omp_set_num_threads()

wrf.omp_get_thread_num

wrf.omp_get_thread_num()
Return the thread number, within the current team, of the calling thread.
The omp_get_thread_num routine returns the thread number of the calling thread, within the team executing the
parallel region to which the routine region binds. The thread number is an integer between 0 and one less than
the value returned by omp_get_num_threads, inclusive. The thread number of the master thread of the team is
0. The routine returns 0 if it is called from the sequential part of a program.

Note: This function always returns 0 when called from within Python.

Returns The thread number.

Return type int

118 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

wrf.omp_set_nested

wrf.omp_set_nested(nested)
Enable or disable nested parallelism, by setting the nest-var ICV
For implementations that support nested parallelism, if the argument to omp_set_nested evaluates to True, nested
parallelism is enabled for the current task; otherwise, nested parallelism is disabled for the current task. For
implementations that do not support nested parallelism, this routine has no effect: the value of nest-var remains
False.
Parameters dynamic_threads (bool) – Set to True to support nested parallelism, otherwise
False.
Returns None
See also:
wrf.omp_get_nested()

wrf.omp_get_nested

wrf.omp_get_nested()
Return the value of the nest-var ICV, which determines if nested parallelism is enabled or disabled
This routine returns 1 if nested parallelism is enabled for the current task; it returns 0, otherwise. If an imple-
mentation does not support nested parallelism, this routine always returns 0.
Returns Returns 1 if nested parallelism is enabled, otherwise 0.
Return type int
See also:
wrf.omp_set_nested()

wrf.omp_set_max_active_levels

wrf.omp_set_max_active_levels(max_levels)
Limit the number of nested active parallel regions on the device, by setting the max-active-levels-var ICV.
The effect of this routine is to set the value of the max-active-levels-var ICV to the value specified in the
argument. If the number of parallel levels requested exceeds the number of levels of parallelism supported
by the implementation, the value of the max-active-levels-var ICV will be set to the number of parallel levels
supported by the implementation. This routine has the described effect only when called from a sequential part
of the program. When called from within an explicit parallel region, the effect of this routine is implementation
defined.
Parameters max_levels (int) – The maximum number of nested active parallel regions.
Returns None.
See also:
wrf.omp_get_max_active_levels()

1.6. API Reference 119

wrf-python Documentation, Release 1.2.0

wrf.omp_get_max_active_levels

wrf.omp_get_max_active_levels()
Return the value of the max-active-levels-var ICV, which determines the maximum number of nested active
parallel regions on the device
The omp_get_max_active_levels routine returns the value of the max-active-levels-var ICV, which determines
the maximum number of nested active parallel regions on the device.
Returns The maximum number of nested active parallel regions.
Return type int
See also:
wrf.omp_set_max_active_levels()

wrf.omp_get_level

wrf.omp_get_level()
Return the value of the levels-var ICV.
The effect of the omp_get_level routine is to return the number of nested parallel regions (whether active or
inactive) enclosing the current task such that all of the parallel regions are enclosed by the outermost initial task
region on the current device.
Returns The number of nested parallel regions.
Return type int
See also:
wrf.omp_get_active_level()

wrf.omp_get_ancestor_thread_num

wrf.omp_get_ancestor_thread_num(level)
Return, for a given nested level of the current thread, the thread number of the ancestor of the current thread.
The omp_get_ancestor_thread_num routine returns the thread number of the ancestor at a given nest level of the
current thread or the thread number of the current thread. If the requested nest level is outside the range of 0 and
the nest level of the current thread, as returned by the omp_get_level routine, the routine returns -1.
Parameters level (int) – The nested level of the current thread.
Returns
The thread number of the ancestor at a given nest level of the current thread.
Return type int
See also:
wrf.omp_get_max_active_levels(), wrf.omp_get_level()

120 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

wrf.omp_get_team_size

wrf.omp_get_team_size(level)
Return, for a given nested level of the current thread, the size of the thread team to which the ancestor or the
current thread belongs
The omp_get_team_size routine returns the size of the thread team to which the ancestor or the current thread
belongs. If the requested nested level is outside the range of 0 and the nested level of the current thread, as
returned by the omp_get_level routine, the routine returns -1. Inactive parallel regions are regarded like active
parallel regions executed with one thread.
Parameters level (int) – The nested level of the current thread.
Returns The size of the thread team.
Return type int
See also:
wrf.omp_get_ancestor_thread_num()

wrf.omp_get_active_level

wrf.omp_get_active_level()
Return the value of the active-level-vars ICV.
The effect of the omp_get_active_level routine is to return the number of nested, active parallel regions enclosing
the current task such that all of the parallel regions are enclosed by the outermost initial task region on the current
device.
Returns The number of nested activate parallel regions.
Return type int
See also:
wrf.omp_get_team_size()

wrf.omp_in_final

wrf.omp_in_final()
Return 1 (True) if the routine is executed in a final task region; otherwise, it returns 0 (False).
Returns Return 1 if the routine is executed in a final task region, 0 otherwise.
Return type int

wrf.omp_init_lock

wrf.omp_init_lock()
Initialize a simple OpenMP lock.
Returns An integer representing the lock.
Return type int
See also:
wrf.omp_init_nest_lock(), wrf.omp_destroy_lock()

1.6. API Reference 121

wrf-python Documentation, Release 1.2.0

wrf.omp_init_nest_lock

wrf.omp_init_nest_lock()
Initialize a nestable OpenMP lock.
Returns An integer representing the nestable lock.
Return type int
See also:
wrf.omp_init_lock()

wrf.omp_destroy_lock

wrf.omp_destroy_lock(svar)
Destroy a simple OpenMP lock.
This sets the lock to an uninitialized state.
Parameters svar (int) – An integer representing the lock.
See also:
wrf.omp_destroy_nest_lock(), wrf.omp_init_lock()

wrf.omp_destroy_nest_lock

wrf.omp_destroy_nest_lock(nvar)
Destroy a nestable OpenMP lock.
This sets the lock to an uninitialized state.
Parameters nvar (int) – An integer representing the nestable lock.
See also:
wrf.omp_destroy_lock(), wrf.omp_init_nest_lock()

wrf.omp_set_lock

wrf.omp_set_lock(svar)
Set a simple OpenMP lock.
Parameters svar (int) – An integer representing the lock.
See also:
wrf.omp_unset_lock(), wrf.omp_set_nest_lock()

wrf.omp_set_nest_lock

wrf.omp_set_nest_lock(nvar)
Set a nestable OpenMP lock.
Parameters nvar (int) – An integer representing the nestable lock.

122 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

See also:
wrf.omp_unset_nest_lock(), wrf.omp_set_lock()

wrf.omp_unset_lock

wrf.omp_unset_lock(svar)
Unset a simple OpenMP lock.
Parameters svar (int) – An integer representing the simple lock.
See also:
wrf.omp_unset_nest_lock(), wrf.omp_set_lock()

wrf.omp_unset_nest_lock

wrf.omp_unset_nest_lock(nvar)
Unset a nestable OpenMP lock.
Parameters nvar (int) – An integer representing the nestable lock.
See also:
wrf.omp_set_nest_lock(), wrf.omp_unset_lock()

wrf.omp_test_lock

wrf.omp_test_lock(svar)
Test a simple OpenMP lock.
This method attempts to set the lock, but does not suspend execution.
Parameters svar (int) – An integer representing the simple lock.
Returns Returns 1 (True) if the lock is successfully set, otherwise 0 (False).
Return type int
See also:
wrf.test_nest_lock(), wrf.omp_set_lock()

wrf.omp_test_nest_lock

wrf.omp_test_nest_lock(nvar)
Test a nestable OpenMP lock.
This method attempts to set the lock, but does not suspend execution.
Parameters nvar (int) – An integer representing the simple lock.
Returns Returns the nesting count if successful, otherwise 0 (False).
Return type int
See also:
wrf.test_lock(), wrf.omp_set_nest_lock()

1.6. API Reference 123

wrf-python Documentation, Release 1.2.0

Configuration Routines

The routines below are used to configure wrf-python by enabling or disabling third party packages. For the most part,
these settings are configured automatically based on the presence of a third party package. However, disabling xarray
can be useful when you want to turn off all metadata in one place.

wrf.xarray_enabled Return True if xarray is installed and enabled.

wrf.disable_xarray Disable xarray.
wrf.enable_xarray Enable xarray.
wrf.cartopy_enabled Return True if cartopy is installed and enabled.
wrf.disable_cartopy Disable cartopy.
wrf.enable_cartopy Enable cartopy.
wrf.basemap_enabled Return True if basemap is installed and enabled.
wrf.disable_basemap Disable basemap.
wrf.enable_basemap Enable basemap.
wrf.pyngl_enabled Return True if pyngl is installed and enabled.
wrf.enable_pyngl Enable pyngl.
wrf.disable_pyngl Disable pyngl.
wrf.set_cache_size Set the maximum number of items that the threadlocal
cache can retain.
wrf.get_cache_size Return the maximum number of items that the threadlo-
cal cache can retain.
wrf.omp_enabled Return True if OpenMP is enabled.

wrf.xarray_enabled

wrf.xarray_enabled()
Return True if xarray is installed and enabled.
Returns True if xarray is installed and enabled.
Return type bool

wrf.disable_xarray

wrf.disable_xarray()
Disable xarray.

wrf.enable_xarray

wrf.enable_xarray()
Enable xarray.

wrf.cartopy_enabled

wrf.cartopy_enabled()
Return True if cartopy is installed and enabled.
Returns True if cartopy is installed and enabled.
Return type bool

124 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

wrf.disable_cartopy

wrf.disable_cartopy()
Disable cartopy.

wrf.enable_cartopy

wrf.enable_cartopy()
Enable cartopy.

wrf.basemap_enabled

wrf.basemap_enabled()
Return True if basemap is installed and enabled.
Returns True if basemap is installed and enabled.
Return type bool

wrf.disable_basemap

wrf.disable_basemap()
Disable basemap.

wrf.enable_basemap

wrf.enable_basemap()
Enable basemap.

wrf.pyngl_enabled

wrf.pyngl_enabled()
Return True if pyngl is installed and enabled.
Returns True if pyngl is installed and enabled.
Return type bool

wrf.enable_pyngl

wrf.enable_pyngl()
Enable pyngl.

wrf.disable_pyngl

wrf.disable_pyngl()
Disable pyngl.

1.6. API Reference 125

wrf-python Documentation, Release 1.2.0

wrf.set_cache_size

wrf.set_cache_size(size)
Set the maximum number of items that the threadlocal cache can retain.
This cache is primarily used for coordinate variables.
Parameters size (int) – The number of items to retain in the cache.
Returns None

wrf.get_cache_size

wrf.get_cache_size()
Return the maximum number of items that the threadlocal cache can retain.
Returns The maximum number of items the cache can retain.
Return type int

Miscellaneous Routines

The routines below are primarily used internally, but some users may find them helpful for other purposes.

wrf.is_time_coord_var Return True if the input variable name is a time coordi-

nate.
wrf.get_coord_pairs Return a tuple for the variable names of the coordi-
nate pair used for the 2D curvilinear coordinate variable.
wrf.is_multi_time_req Return True if the requested time index is for wrf.
ALL_TIMES or None.
wrf.is_multi_file Return True if the input argument is an iterable.
wrf.has_time_coord Return True if the input file or sequence contains the
time coordinate variable.
wrf.is_mapping Return True if the input file or sequence is a mapping
type.
wrf.latlon_coordvars Return the first found latitude and longitude coordinate
names from a NetCDF variable dictionary.
wrf.is_coordvar Returns True if the variable is a coordinate variable.
wrf.get_iterable Returns a resettable iterable object.
wrf.is_moving_domain Return True if the domain is a moving nest.
wrf.npbytes_to_str Return a bytes object for the raw character array.
wrf.is_standard_wrf_var Return True if the variable is a standard WRF variable
and not a diagnostic.
wrf.is_staggered Return True if the variable is on a staggered grid.
wrf.get_left_indexes Returns a tuple for the extra leftmost dimension sizes.
wrf.iter_left_indexes Yield the iteration tuples for a sequence of dimensions
sizes.
wrf.get_right_slices Return an indexing tuple where the left dimensions are
held to a fixed value and the right dimensions are set to
slice objects.
Continued on next page

126 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

Table 14 – continued from previous page

wrf.get_proj_params Return a tuple of latitude, longitude, and projection pa-
rameters from a WRF output file object or a sequence of
WRF output file objects.
wrf.psafilepath Return the full path to the ‘psadilookup.dat’ file.
wrf.get_id Return the cache id.
wrf.getproj Return a wrf.WrfProj subclass.
wrf.cache_item Store an item in the threadlocal cache.
wrf.get_cached_item Return an item from the threadlocal cache.
wrf.ll_points Return the lower left latitude and longitude point(s).
wrf.pairs_to_latlon Return latitude and longitude arrays from a sequence of
wrf.CoordPair objects.

wrf.is_time_coord_var

wrf.is_time_coord_var(varname)
Return True if the input variable name is a time coordinate.
Parameters varname (str) – The input variable name.
Returns True if the input variable is a time coordinate, otherwise False.
Return type bool

wrf.get_coord_pairs

wrf.get_coord_pairs(coord_varname)
Return a tuple for the variable names of the coordinate pair used for the 2D curvilinear coordinate variable.
For example, the ‘XLAT’ variable will have coordinate variables of (‘XLAT’, ‘XLONG’) since the ‘XLAT’
variable itself is two-dimensional.
Parameters coord_varname (str) – The coordinate variable name.
Returns True if the time index is wrf.ALL_TIMES or None, otherwise False.
Return type bool

wrf.is_multi_time_req

wrf.is_multi_time_req(timeidx)
Return True if the requested time index is for wrf.ALL_TIMES or None.
Parameters timeidx (int, wrf.ALL_TIMES, or None) – The requested time index.
Returns True if the time index is wrf.ALL_TIMES or None, otherwise False.
Return type bool

wrf.is_multi_file

wrf.is_multi_file(wrfin)
Return True if the input argument is an iterable.

1.6. API Reference 127

wrf-python Documentation, Release 1.2.0

Parameters wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW

NetCDF data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the afore-
mentioned types.
Returns True if the input is an iterable. False if the input is a single NetCDF file object.
Return type bool

wrf.has_time_coord

wrf.has_time_coord(wrfnc)
Return True if the input file or sequence contains the time coordinate variable.
The time coordinate is named ‘XTIME’.
Parameters wrfnc (netCDF4.Dataset or Nio.NioFile) – A single NetCDF file object.
Returns True if the netcdf file contains the time coordinate variable, False otherwise.
Return type bool

wrf.is_mapping

wrf.is_mapping(wrfin)
Return True if the input file or sequence is a mapping type.
Parameters wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW
NetCDF data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the afore-
mentioned types.
Returns True if the input is a mapping type, False otherwise.
Return type bool

wrf.latlon_coordvars

wrf.latlon_coordvars(ncvars)
Return the first found latitude and longitude coordinate names from a NetCDF variable dictionary.
This function searches the dictionary structure for NetCDF variables and returns the first found latitude and
longitude coordinate names (typically ‘XLAT’ and ‘XLONG’).
Parameters ncvars (dict) – A NetCDF variable dictionary object.
Returns The latitude and longitude coordinate name pairs as (lat_coord_name, lon_coord_name).
Return type tuple

wrf.is_coordvar

wrf.is_coordvar(varname)
Returns True if the variable is a coordinate variable.
Parameters varname (str) – The variable name.
Returns True if the variable is a coordinate variable, otherwise False.
Return type bool

128 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

wrf.get_iterable

wrf.get_iterable(wrfseq)
Returns a resettable iterable object.
In this context, resettable means that when object.__iter__() is called, the iterable returned always
points to the first element in the sequence, similar to how the list and tuple behave.
Parameters wrfseq (iterable) – An iterable type, which includes lists, tuples, dictionaries,
generators, and user-defined classes.
Returns A resettable iterator object.
Return type iterable

wrf.is_moving_domain

wrf.is_moving_domain(wrfin, varname=None, latvar=<wrf.util.either object>, lonvar=<wrf.util.either

object>, _key=None)
Return True if the domain is a moving nest.
This function is used to test for a moving domain, since the WRF output does not set any flags in the file for
this. The test will be performed for all files in any sequences and across all times in each file.
This result is cached internally, so this potentially lengthy check is only done one time for any given wrfin
parameter.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• varname (str, optional) – A specific NetCDF variable to test, which must contain a
‘coordinates’ attribute. If unspecified, The latvar and lonvar parameters are used. Default
is None.
• first_ur_corner (tuple) – A (latitude, longitude) pair for the upper right corner
found in the initial file.
• latvar (str or either, optional) – The latitude variable name. Default is
either('XLAT', 'XLAT_M').
• lonvar (str or either, optional) – The latitude variable name. Default is
either('XLAT', 'XLAT_M').
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
Returns True if the domain is a moving nest, False otherwise.
Return type bool

wrf.npbytes_to_str

wrf.npbytes_to_str(var)
Return a bytes object for the raw character array.
Parameters var (numpy.ndarray) – An array of characters.

1.6. API Reference 129

wrf-python Documentation, Release 1.2.0

Returns A string of bytes.

Return type bytes

wrf.is_standard_wrf_var

wrf.is_standard_wrf_var(wrfin, varname)
Return True if the variable is a standard WRF variable and not a diagnostic.
If wrfin is a sequence, only the first file is used.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• varname (str) – The variable name.
Returns True if the variable is a standard WRF variable, otherwise False.
Return type bool

wrf.is_staggered

wrf.is_staggered(wrfin, var)
Return True if the variable is on a staggered grid.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• var (array) – An array object which contains a shape attribute.
Returns True if the variable is on a staggered grid, otherwise False.
Return type bool

wrf.get_left_indexes

wrf.get_left_indexes(var, expected_dims)
Returns a tuple for the extra leftmost dimension sizes.
For example, if an algorithm expects a 3 dimensional variable, but the variable includes an additional left di-
mension for Time, and this Time dimension has 3 values, then this function will return (3,).
Parameters
• var (array) – An array object that contains the ndim and shape attributes.
• expected_dims (int) – The expected number of dimensions (usually for a computa-
tional algorithm).
Returns The shape for the extra leftmost dimensions.
Return type tuple

130 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

wrf.iter_left_indexes

wrf.iter_left_indexes(dims)
Yield the iteration tuples for a sequence of dimensions sizes.
For example, if dims is (2,2), then this will yield:
(0,0), (0,1), (1,0), (1,1)
This is primarily used to iterate over the leftmost index values.
Parameters dims (indexable sequence) – A sequence of dimension sizes.
Yields tuple – The leftmost indexing iteration sizes.

wrf.get_right_slices

wrf.get_right_slices(var, right_ndims, fixed_val=0)

Return an indexing tuple where the left dimensions are held to a fixed value and the right dimensions are set to
slice objects.
For example, if var is a 5D variable, and the desired indexing sequence for a numpy array is (0,0,0,:,:), then
right_ndims should be set to 2 and fixed_val set to 0.
Parameters
• var (numpy.ndarray) – A numpy array.
• right_ndims (int) – The number of right dimensions to be sliced.
• fixed_val (int) – The value to hold the left dimensions to.
Returns An indexing tuple that can be used to index a numpy.ndarray.
Return type tuple

wrf.get_proj_params

wrf.get_proj_params(wrfin)
Return a tuple of latitude, longitude, and projection parameters from a WRF output file object or a sequence of
WRF output file objects.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. Default is 0.
• varname (str, optional) – The variable name to extract the coordinate variable names
from. Default is None, which will use the default coordinate variable names (‘XLAT’,
‘XLONG’).
Returns A tuple of the latitude coordinate variable, longitude coordinate, and global projection
attributes.
Return type tuple

1.6. API Reference 131

wrf-python Documentation, Release 1.2.0

wrf.psafilepath

wrf.psafilepath()
Return the full path to the ‘psadilookup.dat’ file.
The ‘psadilookup.dat’ file contains the lookup table for the cape routines.
Returns The full path to the ‘psadilookup.dat’ file.
Return type str

wrf.get_id

wrf.get_id(obj, prefix=”)
Return the cache id.
The cache id is used as a caching key for various routines. If the object type is a mapping, then the result will
also be a mapping of each key to the object id for the value.
Parameters
• obj (object) – Any object type.
• prefix (str) – A string to help with recursive calls.
Returns If the obj parameter is not a mapping, then the object id is returned. Otherwise, a mapping
of each key to the object id for the value is returned.
Return type int or dict

wrf.getproj

wrf.getproj(**proj_params)
Return a wrf.WrfProj subclass.
This functions serves as a factory function for returning a wrf.WrfProj subclass from the specified map
projection parameters.
Parameters **proj_params – Map projection optional keyword arguments, that have the same
names as found in WRF output NetCDF global attributes:
• ’MAP_PROJ’: The map projection type as an integer.
• ’TRUELAT1’: True latitude 1.
• ’TRUELAT2’: True latitude 2.
• ’MOAD_CEN_LAT’: Mother of all domains center latitude.
• ’STAND_LON’: Standard longitude.
• ’POLE_LAT’: Pole latitude.
• ’POLE_LON’: Pole longitude.
Returns A wrf.WrfProj subclass for the specified map projection parameters.
Return type wrf.WrfProj

132 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

wrf.cache_item

wrf.cache_item(key, product, value)

Store an item in the threadlocal cache.
The cache should be viewed as two nested dictionaries. The outer key is usually the id for the sequence where
the cached item was generated. The inner key is the product type.
Storing a cached item behaves like:
cache[key][product] = value
The cache is thread local, so stored items are only available in the thread that cached them.
Parameters
• key (int) – The outer dictionary cache key, which is typically the id of the sequence where
the cached item was generated.
• product (str) – The inner dictionary cache key, which is a string for the product type.
• value (object) – The object to store in the cache.
Returns None.
See also:
get_cached_item()

wrf.get_cached_item

wrf.get_cached_item(key, product)
Return an item from the threadlocal cache.
The cache should be viewed as two nested dictionaries. The outer key is usually the id for the sequence where
the cached item was generated. The inner key is the product type.
Retrieving a cached item behaves like:
value = cache[key][product]
The cache is thread local, so stored items are only available in the thread that cached them.
Parameters
• key (int) – The outer dictionary cache key, which is typically the id of the sequence where
the cached item was generated.
• product (str) – The inner dictionary cache key, which is a string for the product type.
Returns The cached object.
Return type object
See also:
cache_item()

wrf.ll_points

wrf.ll_points(lat, lon)
Return the lower left latitude and longitude point(s).

1.6. API Reference 133

wrf-python Documentation, Release 1.2.0

This functions extracts the lower left corner points and returns the result as either a single CoordPair object,
or a list of CoordPair objects.
This is primarily used for testing or constructing the corner point objects from the XLAT and XLONG variables.
Parameters
• lat (xarray.DataArray or numpy.ndarray) – The latitude array. Must be at least
two dimensions.
• lon (xarray.DataArray or numpy.ndarray) – The longitude array. Must be at
least two dimensions.
Returns A single wrf.CoordPair object or a list of wrf.CoordPair objects.
Return type wrf.CoordPair or list

wrf.pairs_to_latlon

wrf.pairs_to_latlon(pairs)
Return latitude and longitude arrays from a sequence of wrf.CoordPair objects.
This function converts a sequence of wrf.CoordPair objects into lists of latitude and longitude points. If
the pairs argument is a single wrf.CoordPair object, then a single latitude and longitude value is returned.
Parameters pairs (wrf.CoordPair or sequence) – A single wrf.CoordPair or sequence
of wrf.CoordPair.
Returns A tuple of (lat, lon), where lat and lon are single values or lists of values.
Return type tuple

Classes

Exceptions

wrf.DiagnosticError Raised when an error occurs in a diagnostic routine.

wrf.DiagnosticError

exception wrf.DiagnosticError(message=None)
Raised when an error occurs in a diagnostic routine.

CoordPair Class

The class below is used for storing coordinate metadata from routines that use a single point for an (x, y) or (lat, lon)
location.

wrf.CoordPair A class that stores (x, y) and/or (latitude, longitude) co-

ordinate pairs.

134 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

wrf.CoordPair

class wrf.CoordPair(x=None, y=None, lat=None, lon=None)

A class that stores (x, y) and/or (latitude, longitude) coordinate pairs.
Most math operators are supplied. When the other operand is a CoordPair, the operation is performed with
the same attribute. When a math operation uses a scalar as the other operand, the operation is applied across all
attributes.
x
float – The x-coordinate.
y
float – The y-coordinate.
lat
float – The latitude coordinate.
lon
float – The longitude coordinate.
__init__(x=None, y=None, lat=None, lon=None)
Initialize a CoordPair object.
Parameters
• x (float, optional) – The x-coordinate.
• y (float, optional) – The y-coordinate.
• lat (float, optional) – The latitude coordinate.
• lon (float, optional) – The longitude coordinate.

Methods

init([x, y, lat, lon]) Initialize a CoordPair object.

latlon_str([fmt]) Return a str for the (latitude, longitude) coordinate
pair.
xy_str([fmt]) Return a str for the (x,y) coordinate pair.

CoordPair Methods

wrf.CoordPair.latlon_str Return a str for the (latitude, longitude) coordinate

pair.
wrf.CoordPair.xy_str Return a str for the (x,y) coordinate pair.

wrf.CoordPair.latlon_str

CoordPair.latlon_str(fmt=’{:.4f}, {:.4f}’)
Return a str for the (latitude, longitude) coordinate pair.
Parameters fmt (str) – The format string. Default is ‘{:.4f}, {:.4f}’
Returns A string for the (latitude, longitude) coordinate pair
Return type str

1.6. API Reference 135

wrf-python Documentation, Release 1.2.0

wrf.CoordPair.xy_str

CoordPair.xy_str(fmt=’{:.4f}, {:.4f}’)
Return a str for the (x,y) coordinate pair.
Parameters fmt (str) – The format string. Default is ‘{:.4f}, {:.4f}’
Returns A string for the (x,y) coordinate pair
Return type str

GeoBounds Class

The class below is used for specifying geographic boundaries.

wrf.GeoBounds A class that stores the geographic boundaries.

wrf.GeoBounds

class wrf.GeoBounds(bottom_left=None, top_right=None, lats=None, lons=None)

A class that stores the geographic boundaries.
Currently, only corner points are used, specified as the bottom left and top right corners. Users can specify the
corner points directly, or specify two-dimensional latitude and longitude arrays and the corner points will be
extracted from them.
bottom_left
wrf.CoordPair – The bottom left coordinate.
top_right
wrf.CoordPair – The top right coordinate.
__init__(bottom_left=None, top_right=None, lats=None, lons=None)
Initialize a wrf.GeoBounds object.
Parameters
• bottom_left (wrf.CoordPair, optional) – The lower left corner. Must also specify
top_right if used. Default is None.
• top_right (wrf.CoordPair, optional) – The upper right corner. Must also specify
bottom_left if used. Default is None.
• lats (numpy.ndarray, optional) – An array of at least two dimensions containing all
of the latitude values. Must also specify lons if used. Default is None.
• lons (numpy.ndarray, optional) – An array of at least two dimensions containing all
of the longitude values. Must also specify lats if used. Default is None.

Methods

init([bottom_left, top_right, lats, lons]) Initialize a wrf.GeoBounds object.

136 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

Projection Classes

The classes below are used to hold the projection information in the ‘projection’ entry within a xarray.
DataArray.attrs attribute.

Projection Base Class

The base class for all map projection types.

wrf.WrfProj A base class for storing map projection information

from WRF data.

wrf.WrfProj

class wrf.WrfProj(**proj_params)
A base class for storing map projection information from WRF data.
Subclasses of this type will be stored in the ‘projection’ attribute entry within a xarray.DataArray.attrs
dictionary. This base class contains the methods required to extract the appropriate projection class for PyNGL,
matplotlib basemap, and cartopy.
map_proj
int – Model projection integer id.
truelat1
float – True latitude 1.
truelat2
float – True latitude 2.
moad_cen_lat
float – Mother of all domains center latitude.
stand_lon
float – Standard longitude.
pole_lat
float – The pole latitude.
pole_lon
float – The pole longitude.
dx
float – The x grid spacing.
dy
float – The y grid spacing.
__init__(**proj_params)
Initialize a wrf.WrfProj object.
Parameters **proj_params – Map projection optional keyword arguments, that have the
same names as found in WRF output NetCDF global attributes (case insensitive):
• ’MAP_PROJ’: The map projection type as an integer.
• ’TRUELAT1’: True latitude 1.

1.6. API Reference 137

wrf-python Documentation, Release 1.2.0

• ’TRUELAT2’: True latitude 2.

• ’MOAD_CEN_LAT’: Mother of all domains center latitude.
• ’STAND_LON’: Standard longitude.
• ’POLE_LAT’: Pole latitude.
• ’POLE_LON’: Pole longitude.

Methods

init(**proj_params) Initialize a wrf.WrfProj object.

basemap(geobounds, **kwargs) Return a matplotlib.mpl_toolkits.
basemap.Basemap object for the map projection.
cartopy() Return a cartopy.crs.Projection subclass
for the map projection.
cartopy_xlim(geobounds) Return the x extents in projected coordinates for car-
topy.
cartopy_ylim(geobounds) Return the y extents in projected coordinates for car-
topy.
cf() Return a dictionary with the NetCDF CF parameters
for the projection.
proj4() Return the PROJ.4 string for the map projection.
pyngl(geobounds, **kwargs) Return a Ngl.Resources object for the map pro-
jection.

Projection Base Class Methods

The class methods for all projection types.

wrf.WrfProj.basemap Return a matplotlib.mpl_toolkits.

basemap.Basemap object for the map projection.
wrf.WrfProj.cartopy Return a cartopy.crs.Projection subclass for
the map projection.
wrf.WrfProj.cartopy_xlim Return the x extents in projected coordinates for car-
topy.
wrf.WrfProj.cartopy_ylim Return the y extents in projected coordinates for car-
topy.
wrf.WrfProj.pyngl Return a Ngl.Resources object for the map projec-
tion.
wrf.WrfProj.cf Return a dictionary with the NetCDF CF parameters for
the projection.
wrf.WrfProj.proj4 Return the PROJ.4 string for the map projection.

wrf.WrfProj.basemap

WrfProj.basemap(geobounds, **kwargs)
Return a matplotlib.mpl_toolkits.basemap.Basemap object for the map projection.
Parameters

138 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

• geobounds (wrf.GeoBounds, optional) – The geobounds to get the extents. If set to

None and using the var parameter, the geobounds will be taken from the variable. If using a
file, then the geobounds will be taken from the native grid.
• **kwargs – Keyword arguments for creating a matplotlib.mpl_toolkits.
basemap.Basemap. By default, the domain bounds will be set to the native projection,
the resolution will be set to ‘l’, and the other projection parameters will be set by the infor-
mation in the file.
Returns A Basemap object for the projection.
Return type matplotlib.mpl_toolkits.basemap.Basemap
See also:
matplotlib.mpl_toolkits.basemap.Basemap

wrf.WrfProj.cartopy

WrfProj.cartopy()
Return a cartopy.crs.Projection subclass for the map projection.
Returns A Projection subclass for the map projection.
Return type cartopy.crs.Projection
See also:
cartopy.crs.Projection

wrf.WrfProj.cartopy_xlim

WrfProj.cartopy_xlim(geobounds)
Return the x extents in projected coordinates for cartopy.
Returns A pair of [xmin, xmax].
Return type list
See also:
cartopy, matplotlib

wrf.WrfProj.cartopy_ylim

WrfProj.cartopy_ylim(geobounds)
Return the y extents in projected coordinates for cartopy.
Returns A pair of [ymin, ymax].
Return type list
See also:
cartopy, matplotlib

1.6. API Reference 139

wrf-python Documentation, Release 1.2.0

wrf.WrfProj.pyngl

WrfProj.pyngl(geobounds, **kwargs)
Return a Ngl.Resources object for the map projection.
Parameters
• geobounds (wrf.GeoBounds, optional) – The geobounds to get the extents. If set to
None and using the var parameter, the geobounds will be taken from the variable. If using a
file, then the geobounds will be taken from the native grid.
• **kwargs – Additional PyNGL resources to set while creating the Ngl.Resources
object.
Returns A dict-like object that contains the PyNGL resources for the map projection.
Return type Ngl.Resources
See also:
PyNGL

wrf.WrfProj.cf

WrfProj.cf()
Return a dictionary with the NetCDF CF parameters for the projection.
Returns:
dict: A dictionary with the NetCDF CF parameter names and projection parameter values.

wrf.WrfProj.proj4

WrfProj.proj4()
Return the PROJ.4 string for the map projection.
Returns A string suitable for use with the PROJ.4 library.
Return type str
See also:
PROJ.4 <https://trac.osgeo.org/proj/>‘_

Projection Subclasses

See wrf.WrfProj for methods and attributes.

wrf.NullProjection A wrf.WrfProj subclass for empty projections.

wrf.LambertConformal A wrf.WrfProj subclass for Lambert Conformal
Conic projections.
wrf.Mercator A wrf.WrfProj subclass for Mercator projections.
wrf.PolarStereographic A wrf.WrfProj subclass for Polar Stereographic
projections.
wrf.LatLon A wrf.WrfProj subclass for Lat Lon projections.
Continued on next page

140 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

Table 24 – continued from previous page

wrf.RotatedLatLon A wrf.WrfProj subclass for Rotated Lat Lon projec-
tions.

wrf.NullProjection

class wrf.NullProjection
A wrf.WrfProj subclass for empty projections.
The NullProjection is primarily used for creating missing projections when using the ‘join’ method.
__init__()
Initialize a wrf.NullProjection object.

Methods

init() Initialize a wrf.NullProjection object.

wrf.LambertConformal

class wrf.LambertConformal(**proj_params)
A wrf.WrfProj subclass for Lambert Conformal Conic projections.
See also:
wrf.WrfProj, wrf.LatLon, wrf.PolarStereographic, Mercator, RotatedLatLon
__init__(**proj_params)
Initialize a wrf.LambertConformal object.
Parameters **proj_params – Map projection optional keyword arguments, that have the
same names as found in WRF output NetCDF global attributes:
• ’TRUELAT1’: True latitude 1.
• ’TRUELAT2’: True latitude 2.
• ’MOAD_CEN_LAT’: Mother of all domains center latitude.
• ’STAND_LON’: Standard longitude.
• ’POLE_LAT’: Pole latitude.

1.6. API Reference 141

wrf-python Documentation, Release 1.2.0

• ’POLE_LON’: Pole longitude.

Methods

init(**proj_params) Initialize a wrf.LambertConformal object.

wrf.Mercator

class wrf.Mercator(**proj_params)
A wrf.WrfProj subclass for Mercator projections.
See also:
wrf.WrfProj, wrf.LatLon, wrf.PolarStereographic, RotatedLatLon,
LambertConformal
__init__(**proj_params)
Initialize a wrf.Mercator object.
Parameters **proj_params – Map projection optional keyword arguments, that have the
same names as found in WRF output NetCDF global attributes:
• ’TRUELAT1’: True latitude 1.
• ’TRUELAT2’: True latitude 2.
• ’MOAD_CEN_LAT’: Mother of all domains center latitude.
• ’STAND_LON’: Standard longitude.
• ’POLE_LAT’: Pole latitude.
• ’POLE_LON’: Pole longitude.

Methods

init(**proj_params) Initialize a wrf.Mercator object.

142 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

Table 27 – continued from previous page

cartopy_xlim(geobounds) Return the x extents in projected coordinates for car-
topy.
cartopy_ylim(geobounds) Return the y extents in projected coordinates for car-
topy.
cf() Return a dictionary with the NetCDF CF parameters
for the projection.
proj4() Return the PROJ.4 string for the map projection.
pyngl(geobounds, **kwargs) Return a Ngl.Resources object for the map pro-
jection.

wrf.PolarStereographic

class wrf.PolarStereographic(**proj_params)
A wrf.WrfProj subclass for Polar Stereographic projections.
See also:
wrf.WrfProj, wrf.LatLon, wrf.RotatedLatLon, Mercator, LambertConformal
__init__(**proj_params)
Initialize a wrf.PolarStereographic object.
Parameters **proj_params – Map projection optional keyword arguments, that have the
same names as found in WRF output NetCDF global attributes:
• ’TRUELAT1’: True latitude 1.
• ’TRUELAT2’: True latitude 2.
• ’MOAD_CEN_LAT’: Mother of all domains center latitude.
• ’STAND_LON’: Standard longitude.
• ’POLE_LAT’: Pole latitude.
• ’POLE_LON’: Pole longitude.

Methods

init(**proj_params) Initialize a wrf.PolarStereographic object.

1.6. API Reference 143

wrf-python Documentation, Release 1.2.0

wrf.LatLon

class wrf.LatLon(**proj_params)
A wrf.WrfProj subclass for Lat Lon projections.
See also:
wrf.WrfProj, wrf.RotatedLatLon, wrf.PolarStereographic, Mercator,
LambertConformal
__init__(**proj_params)
Initialize a wrf.LatLon object.
Parameters **proj_params – Map projection optional keyword arguments, that have the
same names as found in WRF output NetCDF global attributes:
• ’TRUELAT1’: True latitude 1.
• ’TRUELAT2’: True latitude 2.
• ’MOAD_CEN_LAT’: Mother of all domains center latitude.
• ’STAND_LON’: Standard longitude.
• ’POLE_LAT’: Pole latitude.
• ’POLE_LON’: Pole longitude.

Methods

init(**proj_params) Initialize a wrf.LatLon object.

wrf.RotatedLatLon

class wrf.RotatedLatLon(**proj_params)
A wrf.WrfProj subclass for Rotated Lat Lon projections.
See also:
wrf.WrfProj, wrf.LatLon, wrf.PolarStereographic, Mercator, LambertConformal
__init__(**proj_params)
Initialize a wrf.RotatedLatLon object.
Parameters **proj_params – Map projection optional keyword arguments, that have the

144 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

same names as found in WRF output NetCDF global attributes:

• ’TRUELAT1’: True latitude 1.
• ’TRUELAT2’: True latitude 2.
• ’MOAD_CEN_LAT’: Mother of all domains center latitude.
• ’STAND_LON’: Standard longitude.
• ’POLE_LAT’: Pole latitude.
• ’POLE_LON’: Pole longitude.

Methods

init(**proj_params) Initialize a wrf.RotatedLatLon object.

1.6.2 Internal API

Routines

Extraction and Diagnostic Routines

The routines below are called internally by wrf.getvar().

wrf.g_cape.get_2dcape Return the 2d fields of MCAPE, MCIN, LCL, and LFC.

wrf.g_cape.get_3dcape Return the three-dimensional CAPE and CIN.
wrf.g_cloudfrac.get_cloudfrac Return the cloud fraction for low, mid, and high level
clouds.
wrf.g_ctt.get_ctt Return the cloud top temperature.
wrf.g_dbz.get_dbz Return the simulated radar reflectivity.
wrf.g_dbz.get_max_dbz Return the maximum simulated radar reflectivity.
wrf.g_dewpoint.get_dp Return the dewpoint temperature.
wrf.g_dewpoint.get_dp_2m Return the 2m dewpoint temperature.
wrf.g_geoht.get_geopt Return the geopotential.
wrf.g_geoht.get_height Return the geopotential height.
wrf.g_helicity.get_srh Return the storm relative helicity.
wrf.g_helicity.get_uh Return the updraft helicity.
Continued on next page

1.6. API Reference 145

wrf-python Documentation, Release 1.2.0

Table 31 – continued from previous page

wrf.g_omega.get_omega Return Omega.
wrf.g_pressure.get_pressure Return the pressure in the specified units.
wrf.g_pressure.get_pressure_hpa Return the pressure in [hPa].
wrf.g_pw.get_pw Return the preciptiable water.
wrf.g_rh.get_rh Return the relative humidity.
wrf.g_rh.get_rh_2m Return the 2m relative humidity.
wrf.g_slp.get_slp Return the sea level pressure in the specified units.
wrf.g_temp.get_theta Return the potential temperature.
wrf.g_temp.get_temp Return the temperature in the specified units.
wrf.g_temp.get_eth Return the equivalent potential temperature.
wrf.g_temp.get_tv Return the virtual temperature.
wrf.g_temp.get_tw Return the wetbulb temperature.
wrf.g_temp.get_tk Return the temperature in [K].
wrf.g_temp.get_tc Return the temperature in [degC].
wrf.g_terrain.get_terrain Return the terrain height in the specified units.
wrf.g_times.get_times Return a sequence of time objects.
wrf.g_times.get_xtimes Return a sequence of time objects.
wrf.g_uvmet.get_uvmet Return the u,v wind components rotated to earth coor-
dinates.
wrf.g_uvmet.get_uvmet10 Return the u,v components for the 10m winds rotated to
earth coordinates.
wrf.g_uvmet.get_uvmet_wspd_wdir Return the wind speed and wind direction for the wind
rotated to earth coordinates.
wrf.g_uvmet.get_uvmet10_wspd_wdir Return the wind speed and wind direction for the 10m
wind rotated to earth coordinates.
wrf.g_vorticity.get_avo Return the absolute vorticity.
wrf.g_vorticity.get_pvo Return the potential vorticity.
wrf.g_wind.get_u_destag Return the u-component of the wind on mass points.
wrf.g_wind.get_v_destag Return the v-component of the wind on mass points.
wrf.g_wind.get_w_destag Return the w-component of the wind on mass points.
wrf.g_wind.get_destag_wspd_wdir Return the wind speed and wind direction for the wind
in the projected coordinate space.
wrf.g_wind.get_destag_wspd_wdir10 Return the wind speed and wind direction for the 10m
wind in projected coordinate space.

wrf.g_cape.get_2dcape

wrf.g_cape.get_2dcape(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, missing=9.969209968386869e+36)
Return the 2d fields of MCAPE, MCIN, LCL, and LFC.
The leftmost dimension of the returned array represents four different quantities:
• return_val[0,. . . ] will contain MCAPE [J kg-1]
• return_val[1,. . . ] will contain MCIN [J kg-1]
• return_val[2,. . . ] will contain LCL [m]
• return_val[3,. . . ] will contain LFC [m]
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters

146 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF

data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• missing (float) – The fill value to use for the output. Default is wrf.
default_fill(np.float64).
Returns The cape, cin, lcl, and lfc values as an array whose leftmost dimension is 4 (0=CAPE,
1=CIN, 2=LCL, 3=LFC). If xarray is enabled and the meta parameter is True, then the result
will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_cape.get_3dcape

wrf.g_cape.get_3dcape(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, missing=9.969209968386869e+36)
Return the three-dimensional CAPE and CIN.
The leftmost dimension of the returned array represents two different quantities:
• return_val[0,. . . ] will contain CAPE [J kg-1]
• return_val[1,. . . ] will contain CIN [J kg-1]
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.

1.6. API Reference 147

wrf-python Documentation, Release 1.2.0

• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• missing (float) – The fill value to use for the output. Default is wrf.
default_fill(np.float64).
Returns The CAPE and CIN as an array whose leftmost dimension is 2 (0=CAPE, 1=CIN). If xarray
is enabled and the meta parameter is True, then the result will be a xarray.DataArray
object. Otherwise, the result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_cloudfrac.get_cloudfrac

wrf.g_cloudfrac.get_cloudfrac(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None,

meta=True, _key=None, vert_type=’height_agl’,
low_thresh=None, mid_thresh=None, high_thresh=None,
missing=9.969209968386869e+36)
Return the cloud fraction for low, mid, and high level clouds.
The leftmost dimension of the returned array represents three different quantities:
• return_val[0,. . . ] will contain LOW level cloud fraction
• return_val[1,. . . ] will contain MID level cloud fraction
• return_val[2,. . . ] will contain HIGH level cloud fraction
If the vertical coordinate type is ‘height_agl’ or ‘height_msl’, the default cloud levels are defined as:
300 m <= low_cloud < 2000 m 2000 m <= mid_cloud < 6000 m 6000 m <= high_cloud
For ‘pressure’, the default cloud levels are defined as:
97000 Pa <= low_cloud < 80000 Pa 80000 Pa <= mid_cloud < 45000 Pa 45000 Pa <= high_cloud
Note that the default low cloud levels are chosen to exclude clouds near the surface (fog). If you want fog
included, set low_thresh to ~99500 Pa if vert_type is set to ‘pressure’, or 15 m if using ‘height_msl’ or
‘height_agl’. Keep in mind that the lowest mass grid points are slightly above the ground, and in order to
find clouds, the low_thresh needs to be set to values that are slightly greater than (less than) the lowest height
(pressure) values.
When using ‘pressure’ or ‘height_agl’ for vert_type, there is a possibility that the lowest WRF level will be
higher than the low_cloud or mid_cloud threshold, particularly for mountainous regions. When this happens, a
fill value will be used in the output.

148 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• vert_type (str, optional) – The type of vertical coordinate used to determine cloud type
thresholds. Must be ‘height_agl’, ‘height_msl’, or ‘pres’. The default is ‘height_agl’.
• low_thresh (float, optional) – The lower bound for what is considered a low cloud. If
vert_type is ‘pres’, the default is 97000 Pa. If vert_type is ‘height_agl’ or ‘height_msl’, then
the default is 300 m.
• mid_thresh (float, optional) – The lower bound for what is considered a mid level
cloud. If vert_type is ‘pres’, the default is 80000 Pa. If vert_type is ‘height_agl’ or
‘height_msl’, then the default is 2000 m.
• high_thresh (float, optional) – The lower bound for what is considered a high level
cloud. If vert_type is ‘pres’, the default is 45000 Pa. If vert_type is ‘height_agl’ or
‘height_msl’, then the default is 6000 m.
Returns The cloud fraction array whose leftmost dimension is 3 (LOW=0, MID=1, HIGH=2).
If xarray is enabled and the meta parameter is True, then the result will be a xarray.
DataArray object. Otherwise, the result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_ctt.get_ctt

wrf.g_ctt.get_ctt(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, fill_nocloud=False, missing=9.969209968386869e+36,
opt_thresh=1.0, units=’degC’)
Return the cloud top temperature.
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.

1.6. API Reference 149

wrf-python Documentation, Release 1.2.0

Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• fill_nocloud (bool, optional) – Set to True to use fill values in regions where clouds
are not detected (optical depth less than 1). Otherwise, the output will contain the surface
temperature for areas without clouds. Default is False.
• missing (float, optional) – The fill value to use for areas where no clouds are detected.
Only used if fill_nocloud is True. Default is wrf.default_fill(numpy.float64).
• opt_thresh (float, optional) – The amount of optical depth (integrated from top down)
required to trigger a cloud top temperature calculation. The cloud top temperature is cal-
culated at the vertical level where this threshold is met. Vertical columns with less than
this threshold will be treated as cloud free areas. In general, the larger the value is for this
threshold, the lower the altitude will be for the cloud top temperature calculation, and there-
fore higher cloud top temperature values. Default is 1.0, which should be sufficient for most
users.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘ctt’. Default is ‘degC’.
Returns The cloud top temperature. If xarray is enabled and the meta parameter is True, then
the result will be a xarray.DataArray object. Otherwise, the result will be a numpy.
ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_dbz.get_dbz

wrf.g_dbz.get_dbz(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, use_varint=False, use_liqskin=False)
Return the simulated radar reflectivity.
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.

150 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• use_varint (bool, optional) – When set to False, the intercept parameters are assumed
constant (as in MM5’s Reisner-2 bulk microphysical scheme). When set to True, the vari-
able intercept parameters are used as in the more recent version of Reisner-2 (based on
Thompson, Rasmussen, and Manning, 2004, Monthly weather Review, Vol. 132, No. 2, pp.
519-542.).
• use_liqskin (bool, optional) – When set to True, frozen particles that are at a temper-
ature above freezing are assumed to scatter as a liquid particle. Set to False to disable.
Returns The simulated radar reflectivity. If xarray is enabled and the meta parameter is True, then
the result will be a xarray.DataArray object. Otherwise, the result will be a numpy.
ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_dbz.get_max_dbz

wrf.g_dbz.get_max_dbz(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, use_varint=False, use_liqskin=False)
Return the maximum simulated radar reflectivity.
This function returns the maximum reflectivity found in the column for each grid point.
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.

1.6. API Reference 151

wrf-python Documentation, Release 1.2.0

• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• use_varint (bool, optional) – When set to False, the intercept parameters are assumed
constant (as in MM5’s Reisner-2 bulk microphysical scheme). When set to True, the vari-
able intercept parameters are used as in the more recent version of Reisner-2 (based on
Thompson, Rasmussen, and Manning, 2004, Monthly weather Review, Vol. 132, No. 2, pp.
519-542.).
• use_liqskin (bool, optional) – When set to True, frozen particles that are at a temper-
ature above freezing are assumed to scatter as a liquid particle. Set to False to disable.
Returns The maximum simulated radar reflectivity. If xarray is enabled and the meta parameter is
True, then the result will be a xarray.DataArray object. Otherwise, the result will be a
numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_dewpoint.get_dp

wrf.g_dewpoint.get_dp(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, units=’degC’)
Return the dewpoint temperature.
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.

152 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘td’. Default is ‘degC’.
Returns The dewpoint temperature. If xarray is enabled and the meta parameter is True, then the re-
sult will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_dewpoint.get_dp_2m

wrf.g_dewpoint.get_dp_2m(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, units=’degC’)
Return the 2m dewpoint temperature.
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.

1.6. API Reference 153

wrf-python Documentation, Release 1.2.0

• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘td2’. Default is ‘degC’.
Returns The 2m dewpoint temperature. If xarray is enabled and the meta parameter is True, then
the result will be a xarray.DataArray object. Otherwise, the result will be a numpy.
ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_geoht.get_geopt

wrf.g_geoht.get_geopt(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None)
Return the geopotential.
The geopotential is returned in units of [m2 s-2].
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
Returns The geopotential. If xarray is enabled and the meta parameter is True, then the result will
be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray object
with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_geoht.get_height

wrf.g_geoht.get_height(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, msl=True, units=’m’)
Return the geopotential height.

154 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

If msl is True, then geopotential height is returned as Mean Sea Level (MSL). If msl is False, then geopotential
height is returned as Above Ground Level (AGL) by subtracting the terrain height.
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• msl (bool, optional) – Set to True to return geopotential height as Mean Sea Level (MSL).
Set to False to return the geopotential height as Above Ground Level (AGL) by subtracting
the terrain height. Default is True.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘z’. Default is ‘m’.
Returns The geopotential height. If xarray is enabled and the meta parameter is True, then the result
will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_helicity.get_srh

wrf.g_helicity.get_srh(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, top=3000.0)
Return the storm relative helicity.
The top argument specifies the top of the integration in [m].
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.

1.6. API Reference 155

wrf-python Documentation, Release 1.2.0

• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• top (float, optional) – The top of the integration in [m]. Default is 3000.0.
Returns The storm relative helicity. If xarray is enabled and the meta parameter is True, then the re-
sult will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_helicity.get_uh

wrf.g_helicity.get_uh(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, bottom=2000.0, top=5000.0)
Return the updraft helicity.
The bottom and top arguments specify the bottom and top limits for the integration in [m].
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to

156 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• bottom (float, optional) – The bottom limit for the integration in [m]. Default is 2000.0.
• top (float, optional) – The top limit for the integration in [m]. Default is 5000.0.
Returns The updraft helicity. If xarray is enabled and the meta parameter is True, then the result will
be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray object
with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_omega.get_omega

wrf.g_omega.get_omega(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None)
Return Omega.
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
Returns Omega. If xarray is enabled and the meta parameter is True, then the result will be a
xarray.DataArray object. Otherwise, the result will be a numpy.ndarray object with
no metadata.
Return type xarray.DataArray or numpy.ndarray

1.6. API Reference 157

wrf-python Documentation, Release 1.2.0

wrf.g_pressure.get_pressure

wrf.g_pressure.get_pressure(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None,

meta=True, _key=None, units=’Pa’)
Return the pressure in the specified units.
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘pres’. Default is ‘Pa’.
Returns The pressure in the specified units. If xarray is enabled and the meta parameter is True,
then the result will be a xarray.DataArray object. Otherwise, the result will be a numpy.
ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_pressure.get_pressure_hpa

wrf.g_pressure.get_pressure_hpa(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None,

meta=True, _key=None)
Return the pressure in [hPa].
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.

158 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) –
Set to False to disable metadata and return numpy.ndarray instead of xarray.
DataArray. Default is True.
_key (int, optional): A caching key. This is used for internal purposes only. Default is
None.
Returns The pressure in [hPa]. If xarray is enabled and the meta parameter is True, then the result
will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_pw.get_pw

wrf.g_pw.get_pw(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True, _key=None)

Return the preciptiable water.
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.

1.6. API Reference 159

wrf-python Documentation, Release 1.2.0

• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
Returns The preciptable water. If xarray is enabled and the meta parameter is True, then the result
will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_rh.get_rh

wrf.g_rh.get_rh(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True, _key=None)

Return the relative humidity.
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) –
Set to False to disable metadata and return numpy.ndarray instead of xarray.
DataArray. Default is True.
_key (int, optional): A caching key. This is used for internal purposes only. Default is
None.
Returns The relative humidity. If xarray is enabled and the meta parameter is True, then the result
will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

160 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

wrf.g_rh.get_rh_2m

wrf.g_rh.get_rh_2m(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None)
Return the 2m relative humidity.
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
Returns The 2m relative humidity. If xarray is enabled and the meta parameter is True, then the re-
sult will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_slp.get_slp

wrf.g_slp.get_slp(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, units=’hPa’)
Return the sea level pressure in the specified units.
This function extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.

1.6. API Reference 161

wrf-python Documentation, Release 1.2.0

• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘slp’. Default is ‘Pa’.
Returns The sea level pressure in the specified units. If xarray is enabled and the meta parameter
is True, then the result will be a xarray.DataArray object. Otherwise, the result will be a
numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_temp.get_theta

wrf.g_temp.get_theta(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, units=’K’)
Return the potential temperature.
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.

162 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘theta’. Default is ‘K’.
Returns The potential temperature. If xarray is enabled and the meta parameter is True, then the re-
sult will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_temp.get_temp

wrf.g_temp.get_temp(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, units=’K’)
Return the temperature in the specified units.
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘temp’. Default is ‘K’.
Returns The temperature in the specified units. If xarray is enabled and the meta parameter is True,
then the result will be a xarray.DataArray object. Otherwise, the result will be a numpy.
ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

1.6. API Reference 163

wrf-python Documentation, Release 1.2.0

wrf.g_temp.get_eth

wrf.g_temp.get_eth(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, units=’K’)
Return the equivalent potential temperature.
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘eth’. Default is ‘K’.
Returns The equivalent potential temperature. If xarray is enabled and the meta parameter is True,
then the result will be a xarray.DataArray object. Otherwise, the result will be a numpy.
ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_temp.get_tv

wrf.g_temp.get_tv(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, units=’K’)
Return the virtual temperature.
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.

164 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘tv’. Default is ‘K’.
Returns The virtual temperature. If xarray is enabled and the meta parameter is True, then the result
will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_temp.get_tw

wrf.g_temp.get_tw(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, units=’K’)
Return the wetbulb temperature.
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to

1.6. API Reference 165

wrf-python Documentation, Release 1.2.0

repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘twb’. Default is ‘K’.
Returns The wetbulb temperature. If xarray is enabled and the meta parameter is True, then the re-
sult will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_temp.get_tk

wrf.g_temp.get_tk(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None)
Return the temperature in [K].
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
Returns The temperature in [K]. If xarray is enabled and the meta parameter is True, then the result
will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

166 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

wrf.g_temp.get_tc

wrf.g_temp.get_tc(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None)
Return the temperature in [degC].
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
Returns The temperature in [degC]. If xarray is enabled and the meta parameter is True, then the re-
sult will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_terrain.get_terrain

wrf.g_terrain.get_terrain(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None,

meta=False, _key=None, units=’m’)
Return the terrain height in the specified units.
This functions extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.

1.6. API Reference 167

wrf-python Documentation, Release 1.2.0

• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘ter’. Default is ‘m’.
Returns The terrain height in the specified units. If xarray is enabled and the meta parameter is
True, then the result will be a xarray.DataArray object. Otherwise, the result will be a
numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_times.get_times

wrf.g_times.get_times(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None)
Return a sequence of time objects.
This function reads the ‘Times’ variable and creates a sequence of datetime.datetime objects.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES) – The desired time index. This value can be a
positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata.

168 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
Returns A sequence of datetime.datetime objects. If meta is True, the sequence will be of
type xarray.DataArray, otherwise the sequence is numpy.ndarray.
Return type xarray.DataArray or numpy.ndarray

wrf.g_times.get_xtimes

wrf.g_times.get_xtimes(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None)
Return a sequence of time objects.
This function reads the ‘XTIME’ variable and creates a sequence of float objects.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES) – The desired time index. This value can be a
positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
Returns A sequence of float objects. If meta is True, the sequence will be of type xarray.
DataArray, otherwise the sequence is numpy.ndarray.
Return type xarray.DataArray or numpy.ndarray
Raises KeyError – Raised if the ‘XTIME’ variable is not present in the NetCDF file.

wrf.g_uvmet.get_uvmet

wrf.g_uvmet.get_uvmet(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, units=’m s-1’)
Return the u,v wind components rotated to earth coordinates.
The leftmost dimension of the returned array represents two different quantities:
• return_val[0,. . . ] will contain U_EARTH

1.6. API Reference 169

wrf-python Documentation, Release 1.2.0

• return_val[1,. . . ] will contain V_EARTH

This function extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘uvmet’. Default is ‘m s-1’.
Returns The u,v components of the wind rotated to earth coordinates, whose leftmost dimensions
is 2 (0=U_EARTH, 1=V_EARTH). If xarray is enabled and the meta parameter is True, then
the result will be a xarray.DataArray object. Otherwise, the result will be a numpy.
ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_uvmet.get_uvmet10

wrf.g_uvmet.get_uvmet10(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, units=’m s-1’)
Return the u,v components for the 10m winds rotated to earth coordinates.
The leftmost dimension of the returned array represents two different quantities:
• return_val[0,. . . ] will contain U10_EARTH
• return_val[1,. . . ] will contain V10_EARTH
This function extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.

170 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘uvmet10’. Default is ‘m s-1’.
Returns The u,v components of the 10m wind rotated to earth coordinates, whose leftmost dimen-
sions is 2 (0=U10_EARTH, 1=V10_EARTH). If xarray is enabled and the meta parameter is
True, then the result will be a xarray.DataArray object. Otherwise, the result will be a
numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_uvmet.get_uvmet_wspd_wdir

wrf.g_uvmet.get_uvmet_wspd_wdir(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None,

meta=True, _key=None, units=’m s-1’)
Return the wind speed and wind direction for the wind rotated to earth coordinates.
This function should be used when comparing with observations.
The leftmost dimension of the returned array represents two different quantities:
• return_val[0,. . . ] will contain WSPD_EARTH
• return_val[1,. . . ] will contain WDIR_EARTH
This function extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.

1.6. API Reference 171

wrf-python Documentation, Release 1.2.0

• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘uvmet_wspd_wdir’. Default is ‘m s-1’.
Returns The wind speed and wind direction for the wind rotated to earth coordinates, whose left-
most dimensions is 2 (0=WSPD_EARTH, 1=WDIR_EARTH). If xarray is enabled and the meta
parameter is True, then the result will be a xarray.DataArray object. Otherwise, the result
will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_uvmet.get_uvmet10_wspd_wdir

wrf.g_uvmet.get_uvmet10_wspd_wdir(wrfin, timeidx=0, method=’cat’, squeeze=True,

cache=None, meta=True, _key=None, units=’m s-1’)
Return the wind speed and wind direction for the 10m wind rotated to earth coordinates.
This function should be used when comparing with observations.
The leftmost dimension of the returned array represents two different quantities:
• return_val[0,. . . ] will contain WSPD10_EARTH
• return_val[1,. . . ] will contain WDIR10_EARTH
This function extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to

172 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘uvmet10_wspd_wdir’. Default is ‘m s-1’.
Returns The wind speed and wind direction for the 10m wind rotated to earth coordinates, whose
leftmost dimensions is 2 (0=WSPD10_EARTH, 1=WDIR10_EARTH). If xarray is enabled and
the meta parameter is True, then the result will be a xarray.DataArray object. Otherwise,
the result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_vorticity.get_avo

wrf.g_vorticity.get_avo(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None)
Return the absolute vorticity.
This function extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
Returns The absolute vorticity. If xarray is enabled and the meta parameter is True, then the result
will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

1.6. API Reference 173

wrf-python Documentation, Release 1.2.0

wrf.g_vorticity.get_pvo

wrf.g_vorticity.get_pvo(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None)
Return the potential vorticity.
This function extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
Returns The potential vorticity. If xarray is enabled and the meta parameter is True, then the result
will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_wind.get_u_destag

wrf.g_wind.get_u_destag(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, units=’m s-1’)
Return the u-component of the wind on mass points.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.

174 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘wspd_wdir’. Default is ‘m s-1’.
Returns The u-component of the wind. If xarray is enabled and the meta parameter is True, then
the result will be a xarray.DataArray object. Otherwise, the result will be a numpy.
ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_wind.get_v_destag

wrf.g_wind.get_v_destag(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, units=’m s-1’)
Return the v-component of the wind on mass points.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.

1.6. API Reference 175

wrf-python Documentation, Release 1.2.0

• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘wspd_wdir’. Default is ‘m s-1’.
Returns The v-component of the wind. If xarray is enabled and the meta parameter is True, then
the result will be a xarray.DataArray object. Otherwise, the result will be a numpy.
ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.g_wind.get_w_destag

wrf.g_wind.get_w_destag(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None, meta=True,

_key=None, units=’m s-1’)
Return the w-component of the wind on mass points.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘wspd_wdir’. Default is ‘m s-1’.
Returns The w-component of the wind. If xarray is enabled and the meta parameter is True, then
the result will be a xarray.DataArray object. Otherwise, the result will be a numpy.
ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

176 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

wrf.g_wind.get_destag_wspd_wdir

wrf.g_wind.get_destag_wspd_wdir(wrfin, timeidx=0, method=’cat’, squeeze=True, cache=None,

meta=True, _key=None, units=’m s-1’)
Return the wind speed and wind direction for the wind in the projected coordinate space.
The wind speed and direction from this function will be relative to the grid. This function should not be used to
compare with observations, instead use wrf.uvmet10_wspd_wdir() to get the earth relative wind speed
and direction.
The leftmost dimension of the returned array represents two different quantities:
• return_val[0,. . . ] will contain WSPD
• return_val[1,. . . ] will contain WDIR
This function extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘wspd_wdir’. Default is ‘m s-1’.
Returns The wind speed and wind direction for the wind in projected space, whose leftmost dimen-
sions is 2 (0=WSPD, 1=WDIR). If xarray is enabled and the meta parameter is True, then the re-
sult will be a xarray.DataArray object. Otherwise, the result will be a numpy.ndarray
object with no metadata.
Return type xarray.DataArray or numpy.ndarray

1.6. API Reference 177

wrf-python Documentation, Release 1.2.0

wrf.g_wind.get_destag_wspd_wdir10

wrf.g_wind.get_destag_wspd_wdir10(wrfin, timeidx=0, method=’cat’, squeeze=True,

cache=None, meta=True, _key=None, units=’m s-1’)
Return the wind speed and wind direction for the 10m wind in projected coordinate space.
The wind speed and direction from this function will be relative to the grid. This function should not be used to
compare with observations, instead use wrf.uvmet10_wspd_wdir() to get the earth relative wind speed
and direction.
The leftmost dimension of the returned array represents two different quantities:
• return_val[0,. . . ] will contain WSPD10
• return_val[1,. . . ] will contain WDIR10
This function extracts the necessary variables from the NetCDF file object in order to perform the calculation.
Parameters
• wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW NetCDF
data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the aforemen-
tioned types.
• timeidx (int or wrf.ALL_TIMES, optional) – The desired time index. This value can
be a positive integer, negative integer, or wrf.ALL_TIMES (an alias for None) to return all
times in the file or sequence. The default is 0.
• method (str, optional) – The aggregation method to use for sequences. Must be either
‘cat’ or ‘join’. ‘cat’ combines the data along the Time dimension. ‘join’ creates a new
dimension for the file index. The default is ‘cat’.
• squeeze (bool, optional) – Set to False to prevent dimensions with a size of 1 from being
automatically removed from the shape of the output. Default is True.
• cache (dict, optional) – A dictionary of (varname, ndarray) that can be used to supply
pre-extracted NetCDF variables to the computational routines. It is primarily used for in-
ternal purposes, but can also be used to improve performance by eliminating the need to
repeatedly extract the same variables used in multiple diagnostics calculations, particularly
when using large sequences of files. Default is None.
• meta (bool, optional) – Set to False to disable metadata and return numpy.ndarray
instead of xarray.DataArray. Default is True.
• _key (int, optional) – A caching key. This is used for internal purposes only. Default is
None.
• units (str) – The desired units. Refer to the getvar() product table for a list of
available units for ‘wspd_wdir10’. Default is ‘m s-1’.
Returns The wind speed and wind direction for the 10m wind in projected space, whose leftmost
dimensions is 2 (0=WSPD10, 1=WDIR10). If xarray is enabled and the meta parameter is True,
then the result will be a xarray.DataArray object. Otherwise, the result will be a numpy.
ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

178 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

Decorators

Algorithm Decorators

The decorators below are used for performing common operations related to diagnostic and interpolation calculations.

wrf.decorators.convert_units A decorator that converts the units from the wrapped

function’s output.
wrf.decorators.left_iteration A decorator to handle iterating over the leftmost dimen-
sions.
wrf.decorators.cast_type A decorator to handle type casting.
wrf.decorators.extract_and_transpose A decorator to extract the data array from a xarray.
DataArray
wrf.decorators.check_args A decorator to check that the wrapped function’s argu-
ments are valid.
wrf.specialdec.uvmet_left_iter A decorator to handle iterating over the leftmost dimen-
sions for the uvmet diagnostic.
wrf.specialdec.cape_left_iter A decorator to handle iterating over the leftmost dimen-
sions for the cape diagnostic.
wrf.specialdec.cloudfrac_left_iter A decorator to handle iterating over the leftmost dimen-
sions for the cloud fraction diagnostic.

wrf.decorators.convert_units

wrf.decorators.convert_units(unit_type, alg_unit)
A decorator that converts the units from the wrapped function’s output.
The desired units are determined from the wrapped function’s arguments.
Parameters
• unit_type (str) – The unit type. Choices are: ‘wind’, ‘pressure’, ‘temp’, or ‘height’.
• alg_unit (str) – The units returned by the wrapped function, which is usually the units
returned by the Fortran routine.
Returns The wrapped function’s output in the desired units.
Return type numpy.ndarray

wrf.decorators.left_iteration

wrf.decorators.left_iteration(ref_var_expected_dims, ref_var_right_ndims, insert_dims=None,

ref_var_idx=None, ref_var_name=None, ignore_args=None, ig-
nore_kargs=None, outviews=’outview’, alg_dtype=<MagicMock
id=’140087108615304’>, cast_output=True)
A decorator to handle iterating over the leftmost dimensions.
For example, if a wrapped function works with three-dimensional arrays, but the variables include a 4th leftmost
dimension for ‘Time’, this decorator will iterate over all times, call the 3D Fortran routine, and aggregate the
results in to a 4D output array.
It is also important to note that the final output array is allocated first, and then views are passed to the wrapped
function so that values do not need to get copied in to the final output array.

1.6. API Reference 179

wrf-python Documentation, Release 1.2.0

Parameters
• ref_var_expected_dims (int) – The number of dimensions that the Fortran routine
is expecting for the reference variable.
• ref_var_right_ndims (int) – The number of dimensions from the right to keep for
the reference variable when making the output. Can also be a combine_dims object if
the sizes are determined from multiple variables.
• insert_dims (sequence of int, optional) – A sequence of dimensions to insert between
the left dimensions (e.g. time) and the kept right dimensions. Default is None.
• ref_var_idx (int, optional) – The index in the wrapped function’s positional argu-
ments to be used as the reference variable for determining the leftmost dimensions. Must be
specified if ref_var_name is None. Default is None.
• ref_var_name (str, optional) – The keyword argument name for the wrapped func-
tion’s keyword arguments to be used as the reference variable for calculating the leftmost
dimensions. Must be specified if ref_var_idx is None. Default is None.
• ignore_args (sequence of int) – Indexes of any arguments that should be ignored when
creating the sliced views that are passed to the Fortran routine.
• ignore_kargs (sequence of str) – Keys of any keyword arguments that should be ig-
nored when creating the sliced views that are passed to the Fortran routine.
• outviews (str or a sequence) – A single key or sequence of keys that indicate the
wrapped function’s keyword argument to use as the output variable(s) in the wrapped func-
tion.
• alg_dtype (numpy.dtype or str) – The numpy data type used in the wrapped func-
tion.
• cast_output (bool) – Set to True to cast the wrapped function’s output to the same
type as the reference variable.
Returns The aggregated output array that includes all extra leftmost dimensions found in the refer-
ence variable.
Return type numpy.ndarray

wrf.decorators.cast_type

wrf.decorators.cast_type(ref_idx=0, arg_idxs=None, karg_names=None, alg_dtype=<MagicMock

id=’140087108798504’>, outviews=’outview’)
A decorator to handle type casting.
This decorator is used to cast variables to and from the required numpy.dtype used in the wrapped function.
Parameters
• ref_idx (int, optional) – The index in the wrapped function’s positional arguments to
be used as the reference variable for determining the numpy.dtype to return. Default is
0.
• arg_idxs (sequence of int, optional) – A sequence of indexes in the wrapped function’s
positional arguments that indicate which arguments to cast. Must be specified if karg_names
is None. Default is None.
• karg_names (sequence of str) – A sequence of keyword arguments in the wrapped
function’s keyword arguments that indicate the arguments to cast. Must be specified if
arg_idxs is None. Default is None.

180 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

• alg_dtype (numpy.dtype or str) – The numpy data type used in the wrapped func-
tion.
• outviews (str or a sequence) – A single key or sequence of keys that indicate the
wrapped function’s keyword argument to use as the output variable(s) in the wrapped func-
tion.
Returns The wrapped function’s output cast to the same numpy.dtype as the reference variable.
Return type numpy.ndarray

wrf.decorators.extract_and_transpose

wrf.decorators.extract_and_transpose(do_transpose=True, outviews=’outview’)
A decorator to extract the data array from a xarray.DataArray
This decorator also transposes the view of the data to Fortran contiguous if do_transpose is True.
Parameters
• do_transpose – Set to False to only extract the variable. When True, the extracted array
will also be transposed to a Fortran view if it is not already Fortran contiguous.
• outviews (str or a sequence) – A single key or sequence of keys that indicate the
wrapped function’s keyword argument to use as the output variable(s) in the wrapped func-
tion.
Returns A numpy array. If do_transpose is True, the numpy array will also be a Fortran contiguous
view.
Return type numpy.ndarray

wrf.decorators.check_args

wrf.decorators.check_args(refvaridx, refvarndim, rightdims, stagger=None, refstagdim=None)

A decorator to check that the wrapped function’s arguments are valid.
An exception is raised when an invalid argument is found.
Parameters
• refvaridx (int) – The wrapped function’s positional argument index to use as the ref-
erence variable.
• refvarndim (int) – The number of dimensions for the reference variable that is expected
by the wrapped function.
• rightdims (sequence of int) – The expected number of right dimensions for each argu-
ment.
• stagger (sequence of int or None, optional) – The dimension that is staggered for each
argument in the wrapped function. Use None in the sequence to indicate no staggering for
that argument. Default is None.
• refstagdim (int, optional) – The staggered dimension for the reference variable, if
applicable. Default is None.
Returns None
Raises ValueError – Raised when an invalid argument is detected.

1.6. API Reference 181

wrf-python Documentation, Release 1.2.0

wrf.specialdec.uvmet_left_iter

wrf.specialdec.uvmet_left_iter(alg_dtype=<MagicMock id=’140087061208256’>)
A decorator to handle iterating over the leftmost dimensions for the uvmet diagnostic.
For example, if a wrapped function works with three-dimensional arrays, but the variables include a 4th leftmost
dimension for ‘Time’, this decorator will iterate over all times, call the 3D Fortran routine, and aggregate the
results in to a 4D output array.
It is also important to note that the final output array is allocated first, and then views are passed to the wrapped
function so that values do not need to get copied in to the final output array.
Parameters alg_dtype (np.dtype or str) – The numpy data type used in the wrapped func-
tion.
Returns The aggregated uvmet output array that includes all extra leftmost dimensions.
Return type numpy.ndarray

wrf.specialdec.cape_left_iter

wrf.specialdec.cape_left_iter(alg_dtype=<MagicMock id=’140087061222176’>)
A decorator to handle iterating over the leftmost dimensions for the cape diagnostic.
For example, if a wrapped function works with three-dimensional arrays, but the variables include a 4th leftmost
dimension for ‘Time’, this decorator will iterate over all times, call the 3D Fortran routine, and aggregate the
results in to a 4D output array.
It is also important to note that the final output array is allocated first, and then views are passed to the wrapped
function so that values do not need to get copied in to the final output array.
Parameters alg_dtype (np.dtype or str) – The numpy data type used in the wrapped func-
tion.
Returns The aggregated cape output array that includes all extra leftmost dimensions.
Return type numpy.ndarray

wrf.specialdec.cloudfrac_left_iter

wrf.specialdec.cloudfrac_left_iter(alg_dtype=<MagicMock id=’140087061190080’>)
A decorator to handle iterating over the leftmost dimensions for the cloud fraction diagnostic.
For example, if a wrapped function works with three-dimensional arrays, but the variables include a 4th leftmost
dimension for ‘Time’, this decorator will iterate over all times, call the 3D Fortran routine, and aggregate the
results in to a 4D output array.
It is also important to note that the final output array is allocated first, and then views are passed to the wrapped
function so that values do not need to get copied in to the final output array.
Parameters alg_dtype (np.dtype or str) – The numpy data type used in the wrapped func-
tion.
Returns The aggregated cloud fraction output array that includes all extra leftmost dimensions.
Return type numpy.ndarray

182 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

Metadata Decorators

The decorators below are used for performing common operations related to setting metadata.

wrf.metadecorators. A decorator that sets the metadata for a wrapped func-

copy_and_set_metadata tion’s output.
wrf.metadecorators.set_wind_metadata A decorator that sets the metadata for a wrapped wind
function’s output.
wrf.metadecorators.set_cape_metadata A decorator that sets the metadata for a wrapped CAPE
function’s output.
wrf.metadecorators. A decorator that sets the metadata for a wrapped cloud
set_cloudfrac_metadata fraction function’s output.
wrf.metadecorators. A decorator that sets the metadata for a wrapped latlon
set_latlon_metadata function’s output.
wrf.metadecorators. A decorator that sets the metadata for a wrapped height
set_height_metadata function’s output.
wrf.metadecorators. A decorator that sets the metadata for a wrapped inter-
set_interp_metadata polation function.
wrf.metadecorators.set_alg_metadata A decorator that sets the metadata for a wrapped raw
diagnostic function.
wrf.metadecorators. A decorator that sets the metadata for the wrapped raw
set_uvmet_alg_metadata UVMET diagnostic function.
wrf.metadecorators. A decorator that sets the metadata for the wrapped raw
set_cape_alg_metadata CAPE diagnostic function.
wrf.metadecorators. A decorator that sets the metadata for the wrapped raw
set_cloudfrac_alg_metadata cloud fraction diagnostic function.
wrf.metadecorators. A decorator that sets the metadata for the wrapped raw
set_destag_metadata destaggering function.

wrf.metadecorators.copy_and_set_metadata

wrf.metadecorators.copy_and_set_metadata(copy_varname=None, delete_attrs=None,
name=None, remove_dims=None, dim-
names=None, coords=None, **fixed_attrs)
A decorator that sets the metadata for a wrapped function’s output.
Generally, the metadata is copied from the variable specified by copy_varname, with other fixed fields set by
the other decorator arguments.
The cache argument used by most diagnostic routines is supplied by this decorator when the copy_varname
variable is extracted, in order to prevent the variable from being extracted again by the wrapped function.
If the wrapped function’s meta argument is False, then this decorator returns the wrapped function output without
applying the metadata.
Parameters
• copy_varname (str, optional) – The NetCDF variable name to copy. Default is None.
• delete_attrs (sequence of str, optional) – A sequence of key names to remove from
the xarray.DataArray.attrs attribute in the wrapped function output (after being
copied from copy_varname). Default is None.
• name (str) – The name to use for the xarray.DataArray.name attribute.

1.6. API Reference 183

wrf-python Documentation, Release 1.2.0

• remove_dims (sequence of int, optional) – A sequence of dimension indexes to be

removed from the wrapped function output (after being copied from copy_varname). This
is useful when the copy variable is three dimensional but the wrapped function output is two
dimensional, and you still want to keep the names of the rightmost two dimensions. Default
is None.
• dimnames (sequence of str, optional) – A sequence of dimension names in order to
manually set the xarray.DataArray.dims attribute. Default is None.
• coords (dict) – A mapping of coordinate name to coordinate value to manually specify
the xarray.DataArray.coords attribute. Default is None.
• **fixed_attrs – These keyword arguments are added to the xarray.DataArray.
attrs attribute.
Returns The wrapped function output with or without metadata. If xarray is enabled and the meta
parameter is True, then the result will be a xarray.DataArray object. Otherwise, the result
will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.metadecorators.set_wind_metadata

wrf.metadecorators.set_wind_metadata(copy_varname, name, description, wind_ncvar=False,

two_d=False, wspd_wdir=False)
A decorator that sets the metadata for a wrapped wind function’s output.
This is a special metadata decorator for working with wind functions, which include wind extraction routines,
uvmet routines, and wind speed / wind direction routines.
If the wrapped function’s meta argument is False, then this decorator returns the wrapped function output without
applying the metadata.
Parameters
• copy_varname (str, optional) – The NetCDF variable name to copy. Default is None.
• name (str) – The name to use for the xarray.DataArray.name attribute.
• description (str) – The description to use for the ‘description’ key in the xarray.
DataArray.attrs attribute.
• wind_ncvar (bool, optional) – Set to True when the wrapped function is simply extract-
ing a wind variable (U, V, W) from the NetCDF file. Set to False for other types of wind
algorithms (uvmet, wspd_wdir, etc). Default is False.
• two_d (bool, optional) – Set to True if the wind field is two-dimensional. Set to False for
a three-dimensional wind field. Default is False.
• wspd_wdir (bool) – Set to True if the wrapped function is a wind speed / wind direction
algorithm. Otherwise, set to False. Default is False.
Returns The wrapped wind function output with or without metadata. If xarray is enabled and the
meta parameter is True, then the result will be a xarray.DataArray object. Otherwise, the
result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

184 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

wrf.metadecorators.set_cape_metadata

wrf.metadecorators.set_cape_metadata(is2d)
A decorator that sets the metadata for a wrapped CAPE function’s output.
This is a special metadata decorator for working with CAPE functions.
If the wrapped function’s meta argument is False, then this decorator returns the wrapped function output without
applying the metadata.
Parameters is2d (bool) – Set to True if the wrapped function is for a two-dimensional CAPE
routine. Set to False for a three-dimensional CAPE routine.
Returns The wrapped CAPE function output with or without metadata. If xarray is enabled and the
meta parameter is True, then the result will be a xarray.DataArray object. Otherwise, the
result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.metadecorators.set_cloudfrac_metadata

wrf.metadecorators.set_cloudfrac_metadata()
A decorator that sets the metadata for a wrapped cloud fraction function’s output.
This is a special metadata decorator for working with cloud fraction functions.
If the wrapped function’s meta argument is False, then this decorator returns the wrapped function output without
applying the metadata.
Returns The wrapped cloud fraction function output with or without metadata. If xarray is enabled
and the meta parameter is True, then the result will be a xarray.DataArray object. Other-
wise, the result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.metadecorators.set_latlon_metadata

wrf.metadecorators.set_latlon_metadata(xy=False)
A decorator that sets the metadata for a wrapped latlon function’s output.
This is a special metadata decorator for working with latlon functions.
If the wrapped function’s meta argument is False, then this decorator returns the wrapped function output without
applying the metadata.
Parameters xy (bool, optional) – Set to True if the wrapped function returns xy values (ll_to_xy).
Set to False if the wrapped function returns latlon values (xy_to_ll). Default is False.
Returns The wrapped latlon function output with or without metadata. If xarray is enabled and the
meta parameter is True, then the result will be a xarray.DataArray object. Otherwise, the
result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.metadecorators.set_height_metadata

wrf.metadecorators.set_height_metadata(geopt=False, stag=False)
A decorator that sets the metadata for a wrapped height function’s output.

1.6. API Reference 185

wrf-python Documentation, Release 1.2.0

This is a special metadata decorator for working with height functions.

If the wrapped function’s meta argument is False, then this decorator returns the wrapped function output without
applying the metadata.
Parameters
• geopt (bool, optional) – Set to True if the wrapped function returns geopotential. Set to
True if the wrapped function returns geopotential height. Default is False.
• stag (bool, optional) – Set to True to use the vertical staggered grid, rather than the mass
grid. Default is False.
Returns The wrapped height function output with or without metadata. If xarray is enabled and the
meta parameter is True, then the result will be a xarray.DataArray object. Otherwise, the
result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.metadecorators.set_interp_metadata

wrf.metadecorators.set_interp_metadata(interp_type)
A decorator that sets the metadata for a wrapped interpolation function.
If the wrapped function’s meta argument is False, then this decorator returns the wrapped function output without
applying the metadata.
Parameters interp_type (str) – The type of interpolation routine. Choices are: ‘horiz’,
‘cross’, ‘line’, ‘vinterp’, ‘2dxy’, ‘1d’, ‘xy’.
Returns The wrapped interpolation function output with or without metadata. If xarray is enabled
and the meta parameter is True, then the result will be a xarray.DataArray object. Other-
wise, the result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.metadecorators.set_alg_metadata

wrf.metadecorators.set_alg_metadata(alg_ndims, refvarname, refvarndims=None, missin-

garg=None, stagdim=None, stagsubvar=None,
units=None, description=None)
A decorator that sets the metadata for a wrapped raw diagnostic function.
Parameters
• alg_ndims (int) – The number of dimensions returned by the wrapped function.
• refvarname (str) – The wrapped function argument name for the reference variable.
• refvarndims (int, optional) – The number of right dimensions for the reference vari-
able. This paramter is required when the wrapped function result has less dimensions than
reference. Default is None.
• missingarg (str, optional) – The wrapped function argument name for the missing
value variable. Default is None.
• ( (stagdim) – obj‘int‘, optional): The staggered dimension for the reference. This is only
needed if the reference variable is on a staggered grid. Default is None.

186 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

• stagsubvar (str, optional) – The wrapped function argument name to use to supply the
unstaggered dimension name for the staggered dimension in the reference. This is needed if
stagdim is not None. It is primarily used for absolute vorticity. Default is None.
• units (str, optional) – The units to use if if there is no ‘units’ argument for the wrapped
function. Default is None.
• description (str, optional) – A description for the wrapped algorithm, which is stored
in the xarray.DataArray.attrs attribute under the ‘description’ key. Default is
None.
Returns The wrapped numerical function output with or without metadata. If xarray is enabled and
the meta parameter is True, then the result will be a xarray.DataArray object. Otherwise,
the result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.metadecorators.set_uvmet_alg_metadata

wrf.metadecorators.set_uvmet_alg_metadata(units=None, description=’earth rotated u, v’,

latarg=’lat’, windarg=’u’)
A decorator that sets the metadata for the wrapped raw UVMET diagnostic function.
Parameters
• units (str, optional) – The units to use if if there is no ‘units’ argument for the wrapped
function. Default is None.
• description (str, optional) – A description for the wrapped algorithm, which is stored
in the xarray.DataArray.attrs attribute under the ‘description’ key. Default is
None.
• ( (latarg) – obj:’str‘, optional): The wrapped function argument name for latitude. De-
fault is ‘lat’.
• windarg (str, optional) – The wrapped function argument name for the u wind compo-
nent. Default is ‘u’.
Returns The wrapped UVMET function output with or without metadata. If xarray is enabled and
the meta parameter is True, then the result will be a xarray.DataArray object. Otherwise,
the result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.metadecorators.set_cape_alg_metadata

wrf.metadecorators.set_cape_alg_metadata(is2d, copyarg=’pres_hpa’)
A decorator that sets the metadata for the wrapped raw CAPE diagnostic function.
Parameters
• is2d (bool) – Set to True for the two-dimensional CAPE calculation, False for three-
dimensional CAPE.
• copyarg (str) – The wrapped function argument to use for copying dimension names.
Default is ‘pres_hpa’.
Returns The wrapped CAPE function output with or without metadata. If xarray is enabled and the
meta parameter is True, then the result will be a xarray.DataArray object. Otherwise, the
result will be a numpy.ndarray object with no metadata.

1.6. API Reference 187

wrf-python Documentation, Release 1.2.0

Return type xarray.DataArray or numpy.ndarray

wrf.metadecorators.set_cloudfrac_alg_metadata

wrf.metadecorators.set_cloudfrac_alg_metadata(copyarg=’vert’)
A decorator that sets the metadata for the wrapped raw cloud fraction diagnostic function.
Parameters copyarg (str) – The wrapped function argument to use for copying dimension
names. Default is ‘vert’.
Returns The wrapped cloud fraction function output with or without metadata. If xarray is enabled
and the meta parameter is True, then the result will be a xarray.DataArray object. Other-
wise, the result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

wrf.metadecorators.set_destag_metadata

wrf.metadecorators.set_destag_metadata()
A decorator that sets the metadata for the wrapped raw destaggering function.
Returns The wrapped destaggering function output with or without metadata. If xarray is enabled
and the meta parameter is True, then the result will be a xarray.DataArray object. Other-
wise, the result will be a numpy.ndarray object with no metadata.
Return type xarray.DataArray or numpy.ndarray

Decorator Utilities

The routines below are used within the decorators.

wrf.either A callable class that determines which variable is

present in the file.
wrf.combine_dims A callable class that mixes dimension sizes from differ-
ent function arguments.
wrf.from_var A callable class that retrieves attributes from the func-
tion argument.
wrf.from_args Return a mapping of argument name to value for the
called function.
wrf.args_to_list Return all of the function arguments, including defaults,
as a list.
wrf.arg_location Return the function arguments as a single list along with
the index within that list for a specified argument name.

wrf.either

class wrf.either(*varnames)
A callable class that determines which variable is present in the file.
This is used in situations where the same variable type has different names depending on the type of file used.
For example, in a WRF output file, ‘P’ is used for pressure, whereas in a met_em file, pressure is named ‘PRES’.

188 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

__call__(wrfin)
Return the variable that is present in the file.
Parameters wrfin (netCDF4.Dataset, Nio.NioFile, or an iterable) – WRF-ARW
NetCDF data as a netCDF4.Dataset, Nio.NioFile or an iterable sequence of the
aforementioned types.
Returns The variable name that is present in the file.
Return type str
varnames
sequence – A sequence of possible variable names.
__init__(*varnames)
Initialize an either object.
Parameters *varnames (sequence) – A sequence of possible variable names.

Methods

init(*varnames) Initialize an either object.

wrf.combine_dims

class wrf.combine_dims(pairs)
A callable class that mixes dimension sizes from different function arguments.
This callable object is used for determining the output size for the extension module functions. The class is
initialized with a sequence of pairs, where the first value is the function argument index. The second value is
the dimension index(es) to use. The second value can be a single integer or a sequence if multiple dimensions
are used.
__call__(*args)
Return a tuple with the combined dimension sizes.
Parameters *args – The function arguments from which to extract the dimensions sizes.
Returns The shape for the combined dimensions.
Return type tuple
pairs
sequence – A sequence representing how to combine the dimensions.

Example

# Take the -3, and -2 dimension sizes from argument 0

# Then take the -1 dimension size from argument 1
pairs = [(0, (-3, -2), (1, -1)]

combo = combine_dims(pairs)

__init__(pairs)
Initialize a combine_dims object.

1.6. API Reference 189

wrf-python Documentation, Release 1.2.0

Parameters pairs (sequence) – A sequence where each element is a pair (tuple), with the
first element being the function argument index and the second value being either an integer
or sequence for the dimension size indexes to use.

Methods

init(pairs) Initialize a combine_dims object.

wrf.from_var

class wrf.from_var(varname, attribute)

A callable class that retrieves attributes from the function argument.
If the function argument is not of type xarray.DataArray, then None will be returned.
It is assumed that the function has been wrapped using the wrapt module.
__call__(wrapped, *args, **kwargs)
Return the attribute found in the function arguments.
Parameters
• wrapped – The wrapped function, as used by the wrapt module.
• *args – The function arguments.
• **kwargs – The function keyword arguments.
Returns The requested attribute.
Return type object
varname
str – The variable name.
attribute
str – The attribute name.
__init__(varname, attribute)
Initialize a from_var object.
Parameters
• varname (str) – The variable name.
• attribute (str) – The attribute name.

Methods

init(varname, attribute) Initialize a from_var object.

wrf.from_args

wrf.from_args(func, argnames, *args, **kwargs)

Return a mapping of argument name to value for the called function.
This function parses the function *args and **kwargs to obtain the desired argument value. If the argument has

190 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

not been passed in, the value is taken from the default keyword argument value.
This func is usually called from within a decorator.

Note: This function currently does not work with functions that contain *args or **kwargs arguments.

Parameters
• func (function) – The function to examine (usually the function that is wrapped).
• argnames (iterable) – An iterable sequence of argument names.
• *args – The positional arguments.
• **kwargs – The keyword arguments.
Returns A mapping of argument name to argument value.
Return type dict

wrf.args_to_list

wrf.args_to_list(func, args, kwargs)

Return all of the function arguments, including defaults, as a list.
The result can then be passed to the function via result.

Note: This function currently does not work with functions that contain *args or **kwargs arguments.

Parameters
• func (function) – The function to examine (usually the function that is wrapped).
• args (tuple) – The positional arguments.
• kwargs (dict) – The keyword arguments.
Returns A list of all argument values, including defaults.
Return type list

wrf.arg_location

wrf.arg_location(func, argname, args, kwargs)

Return the function arguments as a single list along with the index within that list for a specified argument name.
This function parses the args, kargs and signature looking for the location of argname, and returns a list con-
taining all arguments, along with the argument location in that list.
Parameters
• func (function) – The function to examine (usually the function that is wrapped).
• argname (str) – The argument name to locate.
• args (tuple) – The positional arguments.
• kwargs (dict) – The keyword arguments.

1.6. API Reference 191

wrf-python Documentation, Release 1.2.0

Returns A tuple containing the list of all argument values along with the index for location of
argname.
Return type tuple

Classes

Iterable Wrapper Class

The class below is an Iterable wrapper class and provides an __iter__ function that always returns the beginning of the
sequence, regardless of the Iterable type.

wrf.IterWrapper A wrapper class for generators and custom iterable

classes that returns a new iterator to the start of
the sequence when IterWrapper.__iter__() is
called.

wrf.IterWrapper

class wrf.IterWrapper(wrapped)
A wrapper class for generators and custom iterable classes that returns a new iterator to the start of the sequence
when IterWrapper.__iter__() is called.
If the wrapped object is a generator, a copy of the generator is constructed and returned when IterWrapper.
__iter__() is called. If the wrapped object is a custom type, then the copy.copy() is called and a new
instance is returned. In both cases, the original iterable object is unchanged.

Note: Do not increment the wrapped iterable outside of this wrapper.

__init__(wrapped)
Initialize an wrf.IterWrapper object.
Parameters wrapped (an iterable object) – Any iterable object that contains the
__iter__ method.

Methods

init(wrapped) Initialize an wrf.IterWrapper object.

1.7 Frequently Asked Questions

1.7.1 wrf.getvar() uses more memory when xarray is enabled. Is there a memory
leak?

The difference in memory usage can be observed when running the following code with and without xarray enabled
(see wrf.disable_xarray()):

192 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

from netCDF4 import Dataset

import wrf

# Uncomment to disable xarray

# wrf.disable_xarray()

for i in range(150):
f = Dataset('wrfout_d01_2005-07-17_12_00_00.nc')
p = wrf.getvar(f, 'pressure')
f.close()

When xarray is enabled, there is an internal thread-local cache used to hold the XLAT and XLONG coordinate vari-
ables, along with a boolean variable to remember if a particular file or sequence is from a moving nest. This is useful
when working with sequences of WRF files, so that the XLAT and XLONG variables aren’t repeatedly extracted. This
cache is limited to holding 20 items by default, where the key to each cache entry is the object ID for the file or se-
quence of files. In this particular example, a new file object is created for each iteration, which will have a new object
ID, so the 20 cache spots are going to be filled up. Memory usage will be quite a bit higher for the xarray-enabled
case, but the memory usage will eventually top off.
There is a function wrf.set_cache_size() that can be used to decrease the number of files/sequences in the
cache. By setting the cache size to 0, the cache can be disabled completely.

1.7.2 Can I use xarray.Dataset as an input to the wrf-python functions?

Currently, wrf-python is designed to use the API for PyNIO and netCDF4-python, which differs from the xar-
ray.Dataset API. However, there is an undocumented feature for Dataset objects that can be used as workaround.
Each Dataset object has a xarray.Dataset._file_obj.ds attribute that can be used to obtain the underlying
netCDF file object. Keep in mind that this is not in the public API for xarray, so it could break in a future release.
Here is an example:

import xarray
import wrf

ds = xarray.open_dataset(FILENAME)
slp = wrf.getvar(ds._file_obj.ds, "slp")

In a future release of wrf-python, direct support for Dataset objects will be added and this will no longer be necessary.

1.7.3 Why is wrf-python taking hours to run?

The most likely culprit is insufficient memory for the calculation you are trying to perform.
See Performance Tips for more information.

1.8 Support

1.8.1 Github

For crashes and other issues related to the software, you can submit an issue to the Github repository.
This should be used strictly for crashes and bugs. For general usage questions, please use the Google Group.

1.8. Support 193

wrf-python Documentation, Release 1.2.0

1.8.2 Submitting Files

When issues are encountered, we often need to see the file that is problematic. As long as the file is not so large that it
fills up the FTP server, users can submit files via FTP using the following instructions:
ftp ftp.cgd.ucar.edu
anonymous
<use your email address for the password>
cd incoming
put <your file>
put <your file>
.
.
.
quit

Note: For security reasons, you cannot list the contents of this directory, and neither can we. You must tell us the
exact names of the files that you uploaded in order for us to retrieve them.

1.8.3 Google Group

The best way to receive support is to use the [wrfpython-talk] group on Google. Users can ask usage questions, submit
bug reports, make feature requests, etc. All users of wrf-python are welcome to join.
Below, you can view the most recent discussions:

1.9 Citation

WRF-Python has a Digital Object Identifier (DOI), which is a persistent identifier for web-based resources. The
wrf-python DOI, when used in URL form, https://doi.org/10.5065/D6W094P1, provides a persistent link to the wrf-
python Github page. The benefit of DOIs is that they are widely accepted by academic publishers as citable locators
for scholarly objects.
If you author a paper that involves data analysis with wrf-python, or visualizations created with wrf-python, we would
like to ask you to please cite wrf-python. This helps us better understand the impact of the software on the scientific
community, which in turns helps us maintain support for the effort.
You can cite wrf-python using the following citation:
Ladwig, W. (2017). wrf-python (Version x.x.x) [Software]. Boulder, Colorado: UCAR/
˓→NCAR. https://doi.org/10.5065/D6W094P1

Note: The version number x.x.x should be set to the version of wrf-python that you are using.

1.10 License

PLEASE READ THIS SOFTWARE LICENSE (“LICENSE”) CAREFULLY BEFORE USING THE SOFTWARE.
BY USING THE SOFTWARE, YOU ARE AGREEING TO BE BOUND BY ALL OF THE TERMS OF THIS

194 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

LICENSE. IF YOU DO NOT AGREE TO THE TERMS OF THIS LICENSE, DO NOT USE THE SOFTWARE.
Copyright © 2016 the University Corporation for Atmospheric Research (“UCAR”). All rights reserved. Developed
by NCAR’s Computational and Information Systems Laboratory, UCAR, www2.cisl.ucar.edu.
Redistribution and use of the Software in source and binary forms, with or without modification, is permitted provided
that the following conditions are met:
• Neither the names of NCAR’s Computational and Information Systems Laboratory, the University Corporation
for Atmospheric Research, nor the names of its sponsors or contributors may be used to endorse or promote
products derived from this Software without specific prior written permission.
• Redistributions of source code must retain the above copyright notices, this list of conditions, and the disclaimer
below.
• Redistributions in binary form must reproduce the above copyright notice, this list of conditions, and the dis-
claimer below in the documentation and/or other materials provided with the distribution.
THIS SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED,
INCLUDING, BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PAR-
TICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE CONTRIBUTORS OR COPY-
RIGHT HOLDERS BE LIABLE FOR ANY CLAIM, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
CONSEQUENTIAL DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT
OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
OTHER DEALINGS WITH THE SOFTWARE.

1.11 Tutorials

NCAR occasionally provides tutorials for wrf-python at various times throughout the year.
Below are the links to the upcoming and past tutorials.

1.11.1 Upcoming Tutorials

WRF-Python and VAPOR Workshop 2018 (Boise State University)

The Department of Geosciences at Boise State University is partnering with staff from the National Center for Atmo-
spheric Research (NCAR) to host a free, 2-day workshop in the Environmental Research Building (ERB) lab 2104 at
Boise State University on September 26-27, 2018. The tutorial will be centered on the WRF-Python and VAPOR tools
for analyzing and visualizing data from the Weather Research and Forecasting (WRF) regional weather and climate
model.
Users must be registered to attend this tutorial (see Registration).

Location

September 26-27, 2018 9:00 AM - 4:00 PM

Boise State University, Environmental Research Building (ERB) lab #2104.

WRF-Python Overview

1.11. Tutorials 195

wrf-python Documentation, Release 1.2.0

routines, and utilities to help with plotting via cartopy, basemap, or PyNGL. The functionality is similar to what is
provided by the NCL WRF package.

Note: WRF-Python is NOT a tool for running the WRF-ARW model using Python.

This tutorial provides an introduction to wrf-python. The tutorial is beginner friendly for new users of wrf-python, but
this is NOT an introduction to the Python programming language (see Prerequisites). Due to limited seating, if you
do not have any previous experience with Python, please do not register for this tutorial.

Note: For online training that provides an introduction to the Python programming language itself, please see the
Unidata Python Training Page.

Computers will be provided, but feel free to use your own laptop if you prefer. We will be covering how to install
wrf-python via conda as part of the tutorial.
Students are encouraged to bring their own data sets, but data will be provided if this is not an option. Students will
be provided a jupyter notebook workbook which can be modified to accommodate their data.
Topics include:
• How to install wrf-python via conda
• A brief introduction to jupyter notebook
• Overview of WRF data files
• WRF-Python basics
• Plotting with cartopy
• Overview of OpenMP features and other performance tips
• Open lab for students

Registration

Please register prior to September 19, 2018. The registration form is here:
Registration Form
Registration consists of a brief survey, which will help give the instructor a brief overview of your background and
will help tailor the tutorial to your expectations.

Prerequisites

196 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

• Accessing dict/list/tuple items with the “x[ ]” syntax (e.g. my_list_item = my_list[0]).
• Slicing str/list/tuple with the “:” syntax (e.g. my_slice = my_list[1:3]).
• Using object methods and attributes with the “x.y” syntax (e.g. my_list.append(6)).
• Calling functions (e.g. result = some_function(x, y))
• Familiarity with numpy would be helpful, as only a very brief introduction is provided.
• Familiarity with matplotlib would be helpful, as only a very brief introduction is provided.

Instructions for Computer Lab Installation

Step 1: Download Miniconda

For this tutorial, you will need to download and install Miniconda. We are going to use Python 3.6+.
Please use the appropriate link below to download Miniconda for your operating system.

Note: 64-bit OS recommended

Win64
Mac
Linux
For more information, see: https://conda.io/miniconda.html

Step 2: Install Miniconda

Windows:
1. Browse to the directory where you downloaded Miniconda3-latest-Windows-x86_64.exe.
2. Double click on Miniconda3-latest-Windows-x86_64.exe.
3. Follow the instructions.
4. For Windows 10, use the Anaconda command prompt found under the Anaconda2 menu (Start Menu -> Ana-
conda2 -> Anaconda Prompt). Otherwise, open a regular command prompt.
Mac and Linux:
For Mac and Linux, the installer is a bash script.
1. Using a terminal, you need to execute the bash shell script that you downloaded by doing:

bash /path/to/Miniconda3-latest-MacOSX-x86_64.sh [Mac]

bash /path/to/Miniconda3-latest-Linux-x86_64.sh [Linux]

2. Follow the instructions.

1.11. Tutorials 197

wrf-python Documentation, Release 1.2.0

3. At the end of the installation, it will ask if you want to add the miniconda3 path to your bash
environment. If you are unsure what to do, you should say “yes”. If you say “no”, we’re going to
assume you know what you are doing.
If you said “yes”, then once you restart your shell, the miniconda3 Python will be found instead of
the system Python when you type the “python” command. If you want to undo this later, then you
can edit either ~/.bash_profile or ~/.bashrc (depending on OS used) and comment out the line that
looks similar to:

# added by Miniconda3 x.x.x installer

export PATH="/path/to/miniconda3/bin:$PATH"

4. Restart your command terminal.

5. [Linux and Mac Users Only] Miniconda only works with bash. If bash is not your default shell, then
you need to activate the bash shell by typing the following in to your command terminal:

bash

6. Verify that your system is using the correct Python interpreter by typing the following in to your
command terminal:

which python

You should see the path to your miniconda installation. If not, see the note below.

Note: If you have already installed another Python distribution, like Enthought Canopy, you will
need to comment out any PATH entries for that distribution in your .bashrc or .bash_profile. Other-
wise, your shell environment may pick to wrong Python installation.
If bash is not your default shell type, and the PATH variable has been set in .bash_profile by the
miniconda installer, try executing “bash -l” instead of the “bash” command in step 5.

Step 3: Set Up the Conda Environment

If you are new to the conda package manager, one of the nice features of conda is that you can create isolated Python
environments that prevent package incompatibilities. This is similar to the virtualenv package that some Python users
may be familiar with. However, conda is not compatible with virtualenv, so only use conda environments when
working with conda.
The name of our conda environment for this tutorial is: tutorial_backup.
Follow the instructions below to create the tutorial_backup environment.
1. Open a command terminal if you haven’t done so.
2. [Linux and Mac Users Only] The conda package manager only works with bash, so if bash is not
your current shell, type:

bash

3. Add the conda-forge channel to your conda package manager.

Type or copy the command below in to your command terminal. You should run this command even
if you have already done it in the past. This will ensure that conda-forge is set as the highest priority
channel.

198 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

conda config --add channels conda-forge

Note: Conda-forge is a community driven collection of packages that are continually tested to
ensure compatibility. We highly recommend using conda-forge when working with conda. See
https://conda-forge.github.io/ for more details on this excellent project.

4. Create the backup conda environment for the tutorial.

Students will create a conda environment during the tutorial, but if they run in to problems, we’re
going to create a backup environment.
Type or copy this command in to your command terminal:

conda create -n tutorial_backup python=3.6 matplotlib cartopy netcdf4

˓→jupyter git ffmpeg wrf-python

Type “y” when prompted. It will take several minutes to install everything.
This command creates an isolated Python environment named tutorial_backup, and installs the
python interpreter, matplotlib, cartopy, netcdf4, jupyter, git, ffmpeg, and wrf-python packages.

Note: When the installation completes, your command terminal might post a message similar
to:

If this is your first install of dbus, automatically load on login

˓→with:

mkdir -p ~/Library/LaunchAgents
cp /path/to/miniconda3/envs/tutorial_test/org.freedesktop.dbus-
˓→session.plist ~/Library/LaunchAgents/

launchctl load -w ~/Library/LaunchAgents/org.freedesktop.dbus-

˓→session.plist

This is indicating that the dbus package can be set up to automatically load on login. You can
either ignore this message or type in the commands as indicated on your command terminal.
The tutorial should work fine in either case.

5. Activate the conda environment.

To activate the tutorial_backup Python environment, type the following in to the command terminal:
For Linux and Mac (using bash):

source activate tutorial_backup

For Windows:

activate tutorial_backup

You should see (tutorial_backup) on your command prompt.

To deactivate your conda environment, type the following in to the command terminal:
For Linux and Mac:

1.11. Tutorials 199

wrf-python Documentation, Release 1.2.0

source deactivate

For Windows:

deactivate tutorial_backup

Step 4: Download the Student Workbook

The student workbook for the tutorial is available on GitHub. The tutorial_backup conda environment includes the git
application needed to download the repository.
These instructions download the tutorial in to your home directory. If you want to place the tutorial in to another
directory, we’re going to assume you know how to do this yourself.
To download the student workbook, follow these instructions:
1. Activate the tutorial_backup conda environment following the instructions in the previous step (source activate
tutorial_backup or activate tutorial_backup).
2. Change your working directory to the home directory by typing the following command in to the command
terminal:
For Linux and Mac:

cd ~

For Windows:

cd %HOMEPATH%

3. Download the git repository for the tutorial by typing the following in to the command terminal:

git clone https://github.com/NCAR/wrf_python_tutorial.git

4. There may be additional changes to the tutorial after you have downloaded it. To pull down the latest changes,
type the following in to the command terminal:
For Linux and Mac:

source activate tutorial_backup

cd ~/wrf_python_tutorial/boise_workshop_2018

git pull

For Windows:

activate tutorial_2018

cd %HOMEPATH%\wrf_python_tutorial\boise_workshop_2018

git pull

Note: If you try the “git pull” command and it returns an error indicating that you have made changes to the
workbook, this is probably because you ran the workbook and it contains the cell output. To fix this, first do a
checkout of the workbook, then do the pull.

200 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

git checkout -- .
git pull

Step 5: Verify Your Environment

Verifying that your environment is correct involves importing a few packages and checking for errors (you may see
some warnings for matplotlib or xarray, but you can safely ignore these).
1. Activate the tutorial_backup conda environment if it isn’t already active (see instructions above).
2. Open a python terminal by typing the following in to the command terminal:

python

3. Now type the following in to the Python interpreter:

>>> import netCDF4

>>> import matplotlib
>>> import xarray
>>> import wrf

4. You can exit the Python interpreter using CTRL + D

Step 6: Install WRF Output Files

A link will be provided in an email prior to the tutorial for the WRF-ARW data files used for the examples.
1. The link in the email should take you to a location on an Amazon cloud drive.
2. If you hover your mouse over the wrf_tutorial_data.zip file, you’ll see an empty check box appear next to the
file name. Click this check box.
3. At the bottom of the screen, you’ll see a Download button next to a cloud icon. Click this button to start the
download.
4. The download was most likely placed in to your ~/Downloads folder [%HOMEPATH%\Downloads for Win-
dows]. Using your preferred method of choice for unzipping files, unzip this file in to your home directory. Your
data should now be in ~/wrf_tutorial_data [%HOMEPATH%\wrf_tutorial_data for Windows].
5. Verify that you have three WRF output files in that directory.

1.11.2 Past Tutorials

WRF Users’ Workshop 2017

Welcome wrf-python tutorial attendees!

The instructions below should be completed prior to arriving at the tutorial. There will not be enough time to do this
during the tutorial.

1.11. Tutorials 201

wrf-python Documentation, Release 1.2.0

Prerequisites

This tutorial assumes that you have basic knowledge of how to type commands in to a command terminal using your
preferred operating system. You should know some basic directory commands like cd, mkdir, cp, mv.
Regarding Python, to understand the examples in this tutorial, you should have some experience with Python basics.
Below is a list of some Python concepts that you will see in the examples, but don’t worry if you aren’t familiar with
everything.
• Opening a Python interpreter and entering commands.
• Importing packages via the import statement.
• Familiarity with some of the basic Python types: str, list, tuple, dict, bool, float, int, None.
• Creating a list, tuple, or dict with “[ ]”, “( )”, “{ }” syntax (e.g. my_list = [1,2,3,4,5]).
• Accessing dict/list/tuple items with the “x[ ]” syntax (e.g. my_list_item = my_list[0]).
• Slicing str/list/tuple with the “:” syntax (e.g. my_slice = my_list[1:3]).
• Using object methods and attributes with the “x.y” syntax (e.g. my_list.append(6)).
• Calling functions (e.g. result = some_function(x, y))
• Familiarity with numpy would be helpful, as only a very brief introduction is provided.
• Familiarity with matplotlib would be helpful, as only a very brief introduction is provided.
If you are completely new to Python, that shouldn’t be a problem, since most of the examples consist of basic container
types and function calls. It would be helpful to look at some introductory material before arriving at the tutorial. If
you’ve programmed before, picking up Python isn’t too difficult.
Here are some links:
https://www.learnpython.org/
https://developers.google.com/edu/python/

Step 1: Open a Command Terminal

To begin, you will first need to know how to open a command line terminal for your operating system.
For Windows:

WINDOWS + r
type cmd in the run window

For Mac:

Finder -> Applications -> Utilities -> Terminal

For Linux:

Try one of the following:

CTRL + ALT + T
CTRL + ALT + F2

202 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

Step 2: Download Miniconda

For this tutorial, you will need to download and install Miniconda. We are going to use Python 2.7, but it should also
work with Python 3.5+. However, due to limitations with open source compilers on conda-forge, only Python 2.7 is
available for Windows.
Please use the appropriate link below to download Miniconda for your operating system.

Note: 64-bit OS recommended

Win64
Mac
Linux
For more information, see: https://conda.io/miniconda.html

Note: What is Miniconda?

If you have used the Anaconda distribution for Python before, then you will be familiar with Miniconda. The Anaconda
Python distribution includes numerous scientific packages out of the box, which can be difficult for users to build and
install. More importantly, Anaconda includes the conda package manager.
The conda package manager is a utility (similar to yum or apt-get) that installs packages from a repository of pre-
compiled Python packages. These repositories are called channels. Conda makes it easy for Python users to install
and uninstall packages, and also can be used to create isolated Python environments (more on that later).
Miniconda is a bare bones implementation of Anaconda and only includes the conda package manager. Since we are
going to use the conda-forge channel to install our scientific packages, Miniconda avoids any complications between
packages provided by Anaconda and conda-forge.

Step 3: Install Miniconda

Windows:
1. Browse to the directory where you downloaded Miniconda2-latest-Windows-x86_64.exe.
2. Double click on Miniconda2-latest-Windows-x86_64.exe.
3. Follow the instructions.
4. Restart your command terminal.
Mac and Linux:
For Mac and Linux, the installer is a bash script.
1. Using a terminal, you need to execute the bash shell script that you downloaded by doing:

bash /path/to/Miniconda2-latest-MacOSX-x86_64.sh [Mac]

bash /path/to/Miniconda2-latest-Linux-x86_64.sh [Linux]

2. Follow the instructions.

1.11. Tutorials 203

wrf-python Documentation, Release 1.2.0

3. At the end of the installation, it will ask if you want to add the miniconda2 path to your bash
environment. If you are unsure what to do, you should say “yes”. If you say “no”, we’re going to
assume you know what you are doing.
If you said “yes”, then once you restart your shell, the miniconda2 Python will be found instead of
the system Python when you type the “python” command. If you want to undo this later, then you
can edit either ~/.bash_profile or ~/.bashrc (depending on OS used) and comment out the line that
looks similar to:

# added by Miniconda2 4.1.11 installer

export PATH="/path/to/miniconda2/bin:$PATH"

4. Restart your command terminal.

5. [Linux and Mac Users Only] Miniconda only works with bash. If bash is not your default shell, then
you need to activate the bash shell by typing the following in to your command terminal:

bash

6. Verify that your system is using the correct Python interpreter by typing the following in to your
command terminal:

which python

You should see the path to your miniconda installation. If not, see the note below.

Step 4: Set Up the Conda Environment

If you are new to the conda package manager, one of the nice features of conda is that you can create isolated Python
environments that prevent package incompatibilities. This is similar to the virtualenv package that some Python users
may be familiar with. However, conda is not compatible with virtualenv, so only use conda environments when
working with conda.
The name of our conda environment for this tutorial is: tutorial_2017.
Follow the instructions below to create the tutorial_2017 environment.
1. Open a command terminal if you haven’t done so.
2. [Linux and Mac Users Only] The conda package manager only works with bash, so if bash is not
your current shell, type:

bash

3. Add the conda-forge channel to your conda package manager.

204 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

conda config --add channels conda-forge

4. Create the conda environment for the tutorial.

Type or copy this command in to your command terminal:

conda create -n tutorial_2017 python=2.7 matplotlib=1.5.3 cartopy

˓→netcdf4 jupyter git ffmpeg wrf-python

Type “y” when prompted. It will take several minutes to install everything.
This command creates an isolated Python environment named tutorial_2017, and installs the python
interpreter, matplotlib, cartopy, netcdf4, jupyter, git, ffmpeg, and wrf-python packages.

Note: When the installation completes, your command terminal might post a message similar
to:

If this is your first install of dbus, automatically load on login

˓→with:

mkdir -p ~/Library/LaunchAgents
cp /path/to/miniconda2/envs/tutorial_test/org.freedesktop.dbus-
˓→session.plist ~/Library/LaunchAgents/

launchctl load -w ~/Library/LaunchAgents/org.freedesktop.dbus-

˓→session.plist

Note: In this tutorial, we need to use matplotlib v1.5.3 due to some issues with cartopy, which
should be fixed in a later version of cartopy. Be sure to supply the version number as indicated
in the command above.

5. Activate the conda environment.

To activate the tutorial_2017 Python environment, type the following in to the command terminal:
For Linux and Mac (using bash):

source activate tutorial_2017

For Windows:

activate tutorial_2017

You should see (tutorial_2017) on your command prompt.

To deactivate your conda environment, type the following in to the command terminal:

1.11. Tutorials 205

wrf-python Documentation, Release 1.2.0

For Linux and Mac:

source deactivate

For Windows:
deactivate tutorial_2017

Step 5: Download the Student Workbook

The student workbook for the tutorial is available on GitHub. The tutorial_2017 conda environment includes the git
application needed to download the repository.
These instructions download the tutorial in to your home directory. If you want to place the tutorial in to another
directory, we’re going to assume you know how to do this yourself.
To download the student workbook, follow these instructions:
1. Activate the tutorial_2017 conda environment following the instructions in the previous step (source activate
tutorial_2017 or activate tutorial_2017).
2. Change your working directory to the home directory by typing the following command in to the command
terminal:
For Linux and Mac:
cd ~

For Windows:
cd %HOMEPATH%

3. Download the git repository for the tutorial by typing the following in to the command terminal:
git clone https://github.com/NCAR/wrf_python_tutorial.git

4. There may be additional changes to the tutorial after you have downloaded it. To pull down the latest changes,
type the following in to the command terminal:
For Linux and Mac:
source activate tutorial_2017

cd ~/wrf_python_tutorial

git pull

For Windows:
activate tutorial_2017

cd %HOMEPATH%\wrf_python_tutorial

git pull

206 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

git checkout -- wrf_workshop_2017.ipynb

git pull

Step 6: Verify Your Environment

Verifying that your environment is correct involves importing a few packages and checking for errors (you may see
some warnings for matplotlib or xarray, but you can safely ignore these).
1. Activate the tutorial_2017 conda environment if it isn’t already active (see instructions above).
2. Open a python terminal by typing the following in to the command terminal:

python

3. Now type the following in to the Python interpreter:

>>> import netCDF4

>>> import matplotlib
>>> import xarray
>>> import wrf

4. You can exit the Python interpreter using CTRL + D

Step 7: Obtain WRF Output Files

For this tutorial, we strongly recommend that you use your own WRF output files. The tutorial includes an easy way
to point to your own data files. The WRF output files should all be from the same WRF run and use the same domain.
If your files are located on another system (e.g. yellowstone), then copy 2 or 3 of these files to your local computer
prior to the tutorial.
If you do not have any of your own WRF output files, then you can download the instructor data files from a link that
should have been provided to you in an email prior to the tutorial.
If you are using the link provided in the email for your data, you can follow the instructions below to place your data
in the default location for your workbook.
1. The link in the email should take you to a location on an Amazon cloud drive.
2. If you hover your mouse over the wrf_tutorial_data.zip file, you’ll see an empty check box appear next to the
file name. Click this check box.
3. At the bottom of the screen, you’ll see a Download button next to a cloud icon. Click this button to start the
download.
4. The download was most likely placed in to your ~/Downloads folder [%HOMEPATH%\Downloads for Win-
dows]. Using your preferred method of choice for unzipping files, unzip this file in to your home directory. Your
data should now be in ~/wrf_tutorial_data [%HOMEPATH%\wrf_tutorial_data for Windows].
5. Verify that you have three WRF output files in that directory.

Getting Help

If you experience problems during this installation, please send a question to the Google Group support mailing list.
We look forward to seeing you at the tutorial!

1.11. Tutorials 207

wrf-python Documentation, Release 1.2.0

WRF-Python Tutorial 2018

NCAR will be providing a four hour tutorial for wrf-python on Wednesday, March 7, 2018 from 8 AM to 12 PM. The
tutorial is free, but seating is limited to only 16 students, so registration is required.
The tutorial will take place at NCAR’s corporate training center in Boulder, Colorado.
Corporate Technical Training Center 3085 Center Green Drive, Building CG-2, Room #3024 Boulder, Colorado

Overview

WRF-Python is a collection of diagnostic and interpolation routines for use with output from the Weather Research
and Forecasting (WRF-ARW) Model. The package provides over 30 diagnostic calculations, several interpolation
routines, and utilities to help with plotting via cartopy, basemap, or PyNGL. The functionality is similar to what is
provided by the NCL WRF package.

Note: WRF-Python is NOT a tool for running the WRF-ARW model using Python.

Note: For online training that provides an introduction to the Python programming language itself, please see the
Unidata Python Training Page.

Registration

Please register prior to February 21, 2018. The registration form is here:
Registration Form
Registration consists of a brief survey, which will help give the instructor a brief overview of your background and
will help tailor the tutorial to your expectations.

208 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

Prerequisites

This tutorial assumes that you have basic knowledge of how to type commands in to a terminal window using your
preferred operating system. You should know some basic directory commands like cd, mkdir, cp, mv.
This tutorial assumes that you have prior experience programming in Python. Below is a list of some Python concepts
that you will see in the examples, but don’t worry if you aren’t familiar with everything.
• Opening a Python interpreter and entering commands.
• Importing packages via the import statement.
• Familiarity with some of the basic Python types: str, list, tuple, dict, bool, float, int, None.
• Creating a list, tuple, or dict with “[ ]”, “( )”, “{ }” syntax (e.g. my_list = [1,2,3,4,5]).
• Accessing dict/list/tuple items with the “x[ ]” syntax (e.g. my_list_item = my_list[0]).
• Slicing str/list/tuple with the “:” syntax (e.g. my_slice = my_list[1:3]).
• Using object methods and attributes with the “x.y” syntax (e.g. my_list.append(6)).
• Calling functions (e.g. result = some_function(x, y))
• Familiarity with numpy would be helpful, as only a very brief introduction is provided.
• Familiarity with matplotlib would be helpful, as only a very brief introduction is provided.

WRF Users’ Workshop 2018

Welcome WRF-Python Tutorial attendees!

If you wish to actively participate in the tutorial, please bring your own laptop. Due to limited time constraints, the
instructions below should be completed prior to arriving at the tutorial.
I will be executing the same cells as the student workbook, so if you prefer to sit and watch, that’s OK too. Following
the tutorial, I will upload the instructor slides in to the same GitHub location as the student workbook if you want to
try it out later.

Prerequisites

1.11. Tutorials 209

wrf-python Documentation, Release 1.2.0

• Familiarity with numpy would be helpful, as only a very brief introduction is provided.
• Familiarity with matplotlib would be helpful, as only a very brief introduction is provided.
If you are completely new to Python, that shouldn’t be a problem, since most of the examples consist of basic container
types and function calls. It would be helpful to look at some introductory material before arriving at the tutorial. If
you’ve programmed before, picking up Python isn’t too difficult.
Here are some links:
https://www.learnpython.org/
https://developers.google.com/edu/python/

Step 1: Open a Command Terminal

To begin, you will first need to know how to open a command line terminal for your operating system.
For Windows:

WINDOWS + r
type cmd in the run window

For Mac:

Finder -> Applications -> Utilities -> Terminal

For Linux:

Try one of the following:

CTRL + ALT + T
CTRL + ALT + F2

Step 2: Download Miniconda

For this tutorial, you will need to download and install Miniconda. We are going to use Python 3.6, but it will also
work with Python 2.7.
Please use the appropriate link below to download Miniconda for your operating system.

Note: 64-bit OS recommended

Win64
Mac
Linux
For more information, see: https://conda.io/miniconda.html

Note: What is Miniconda?

210 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

The conda package manager is a utility (similar to yum or apt-get) that installs packages from a repository of pre-
compiled Python packages. These repositories are called channels. Conda makes it easy for Python users to install
and uninstall packages, and also can be used to create isolated Python environments (more on that later).
Miniconda is a bare bones implementation of Anaconda and only includes the conda package manager. Since we are
going to use the conda-forge channel to install our scientific packages, Miniconda avoids any complications between
packages provided by Anaconda and conda-forge.

Step 3: Install Miniconda

Windows:
1. Browse to the directory where you downloaded Miniconda3-latest-Windows-x86_64.exe.
2. Double click on Miniconda3-latest-Windows-x86_64.exe.
3. Follow the instructions.
4. Restart your command terminal.
Mac and Linux:
For Mac and Linux, the installer is a bash script.
1. Using a terminal, you need to execute the bash shell script that you downloaded by doing:

bash /path/to/Miniconda3-latest-MacOSX-x86_64.sh [Mac]

bash /path/to/Miniconda3-latest-Linux-x86_64.sh [Linux]

2. Follow the instructions.

# added by Miniconda3 x.x.x installer

export PATH="/path/to/miniconda3/bin:$PATH"

4. Restart your command terminal.

5. [Linux and Mac Users Only] Miniconda only works with bash. If bash is not your default shell, then
you need to activate the bash shell by typing the following in to your command terminal:

bash

6. Verify that your system is using the correct Python interpreter by typing the following in to your
command terminal:

which python

You should see the path to your miniconda installation. If not, see the note below.

1.11. Tutorials 211

wrf-python Documentation, Release 1.2.0

Step 4: Set Up the Conda Environment

If you are new to the conda package manager, one of the nice features of conda is that you can create isolated Python
environments that prevent package incompatibilities. This is similar to the virtualenv package that some Python users
may be familiar with. However, conda is not compatible with virtualenv, so only use conda environments when
working with conda.
The name of our conda environment for this tutorial is: tutorial_2018.
Follow the instructions below to create the tutorial_2018 environment.
1. Open a command terminal if you haven’t done so.
2. [Linux and Mac Users Only] The conda package manager only works with bash, so if bash is not
your current shell, type:

bash

3. Add the conda-forge channel to your conda package manager.

conda config --add channels conda-forge

4. Create the conda environment for the tutorial.

Type or copy this command in to your command terminal:

conda create -n tutorial_2018 python=3.6 matplotlib cartopy netcdf4

˓→jupyter git ffmpeg wrf-python

Type “y” when prompted. It will take several minutes to install everything.
This command creates an isolated Python environment named tutorial_2018, and installs the python
interpreter, matplotlib, cartopy, netcdf4, jupyter, git, ffmpeg, and wrf-python packages.

Note: When the installation completes, your command terminal might post a message similar
to:

212 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

If this is your first install of dbus, automatically load on login

˓→with:

mkdir -p ~/Library/LaunchAgents
cp /path/to/miniconda3/envs/tutorial_test/org.freedesktop.dbus-
˓→session.plist ~/Library/LaunchAgents/

launchctl load -w ~/Library/LaunchAgents/org.freedesktop.dbus-

˓→session.plist

5. Activate the conda environment.

To activate the tutorial_2018 Python environment, type the following in to the command terminal:
For Linux and Mac (using bash):

source activate tutorial_2018

For Windows:

activate tutorial_2018

You should see (tutorial_2018) on your command prompt.

To deactivate your conda environment, type the following in to the command terminal:
For Linux and Mac:

source deactivate

For Windows:

deactivate tutorial_2018

Step 5: Download the Student Workbook

The student workbook for the tutorial is available on GitHub. The tutorial_2018 conda environment includes the git
application needed to download the repository.
These instructions download the tutorial in to your home directory. If you want to place the tutorial in to another
directory, we’re going to assume you know how to do this yourself.
To download the student workbook, follow these instructions:
1. Activate the tutorial_2018 conda environment following the instructions in the previous step (source activate
tutorial_2018 or activate tutorial_2018).
2. Change your working directory to the home directory by typing the following command in to the command
terminal:
For Linux and Mac:

cd ~

For Windows:

1.11. Tutorials 213

wrf-python Documentation, Release 1.2.0

cd %HOMEPATH%

3. Download the git repository for the tutorial by typing the following in to the command terminal:

git clone https://github.com/NCAR/wrf_python_tutorial.git

4. There may be additional changes to the tutorial after you have downloaded it. To pull down the latest changes,
type the following in to the command terminal:
For Linux and Mac:

source activate tutorial_2018

cd ~/wrf_python_tutorial/wrf_workshop_2018

git pull

For Windows:

activate tutorial_2018

cd %HOMEPATH%\wrf_python_tutorial\wrf_workshop_2018

git pull

git checkout -- .
git pull

Step 6: Verify Your Environment

Verifying that your environment is correct involves importing a few packages and checking for errors (you may see
some warnings for matplotlib or xarray, but you can safely ignore these).
1. Activate the tutorial_2018 conda environment if it isn’t already active (see instructions above).
2. Open a python terminal by typing the following in to the command terminal:

python

3. Now type the following in to the Python interpreter:

>>> import netCDF4

>>> import matplotlib
>>> import xarray
>>> import wrf

4. You can exit the Python interpreter using CTRL + D

214 Chapter 1. Documentation

wrf-python Documentation, Release 1.2.0

Step 7: Obtain WRF Output Files

A link will be provided in an email prior to the tutorial for the WRF-ARW data files used for the examples. If you did
not receive this email, the link will also be provided at the tutorial itself.
You also have the option of using your own data files for the tutorial by modifying the first Jupyter Notebook cell
to point to your data set. However, there is no guarantee that every cell in your workbook will work without some
modifications (e.g. cross section lines will be drawn outside of your domain).
1. The link in the email should take you to a location on an Amazon cloud drive.
2. If you hover your mouse over the wrf_tutorial_data.zip file, you’ll see an empty check box appear next to the
file name. Click this check box.
3. At the bottom of the screen, you’ll see a Download button next to a cloud icon. Click this button to start the
download.
4. The download was most likely placed in to your ~/Downloads folder [%HOMEPATH%\Downloads for Win-
dows]. Using your preferred method of choice for unzipping files, unzip this file in to your home directory. Your
data should now be in ~/wrf_tutorial_data [%HOMEPATH%\wrf_tutorial_data for Windows].
5. Verify that you have three WRF output files in that directory.

Getting Help

If you experience problems during this installation, please send a question to the Google Group support mailing list.
We look forward to seeing you at the tutorial!

1.11. Tutorials 215

wrf-python Documentation, Release 1.2.0

216 Chapter 1. Documentation

CHAPTER 2

Indices and tables

• genindex
• modindex
• search

217
wrf-python Documentation, Release 1.2.0

218 Chapter 2. Indices and tables

Index

Symbols cartopy_ylim() (wrf.WrfProj method), 139

__call__() (wrf.combine_dims method), 189 cast_type() (in module wrf.decorators), 180
__call__() (wrf.either method), 188 cf() (wrf.WrfProj method), 140
__call__() (wrf.from_var method), 190 check_args() (in module wrf.decorators), 181
__init__() (wrf.CoordPair method), 135 cloudfrac() (in module wrf), 101
__init__() (wrf.GeoBounds method), 136 cloudfrac_left_iter() (in module wrf.specialdec), 182
__init__() (wrf.IterWrapper method), 192 combine_dims (class in wrf), 189
__init__() (wrf.LambertConformal method), 141 combine_files() (in module wrf), 79
__init__() (wrf.LatLon method), 144 convert_units() (in module wrf.decorators), 179
__init__() (wrf.Mercator method), 142 CoordPair (class in wrf), 135
__init__() (wrf.NullProjection method), 141 copy_and_set_metadata() (in module
__init__() (wrf.PolarStereographic method), 143 wrf.metadecorators), 183
__init__() (wrf.RotatedLatLon method), 144 ctt() (in module wrf), 102
__init__() (wrf.WrfProj method), 137
__init__() (wrf.combine_dims method), 189
D
__init__() (wrf.either method), 189 dbz() (in module wrf), 104
__init__() (wrf.from_var method), 190 destagger() (in module wrf), 78
DiagnosticError, 134
A disable_basemap() (in module wrf), 125
disable_cartopy() (in module wrf), 125
arg_location() (in module wrf), 191
disable_pyngl() (in module wrf), 125
args_to_list() (in module wrf), 191
disable_xarray() (in module wrf), 124
attribute (wrf.from_var attribute), 190
dx (wrf.WrfProj attribute), 137
avo() (in module wrf), 107
dy (wrf.WrfProj attribute), 137
B E
basemap() (wrf.WrfProj method), 138
either (class in wrf), 188
basemap_enabled() (in module wrf), 125
enable_basemap() (in module wrf), 125
bottom_left (wrf.GeoBounds attribute), 136
enable_cartopy() (in module wrf), 125
enable_pyngl() (in module wrf), 125
C enable_xarray() (in module wrf), 124
cache_item() (in module wrf), 133 eth() (in module wrf), 109
cape_2d() (in module wrf), 98 extract_and_transpose() (in module wrf.decorators), 181
cape_3d() (in module wrf), 100 extract_dim() (in module wrf), 80
cape_left_iter() (in module wrf.specialdec), 182 extract_global_attrs() (in module wrf), 81
cartopy() (wrf.WrfProj method), 139 extract_times() (in module wrf), 81
cartopy_enabled() (in module wrf), 124 extract_vars() (in module wrf), 79
cartopy_xlim() (in module wrf), 86
cartopy_xlim() (wrf.WrfProj method), 139 F
cartopy_ylim() (in module wrf), 87 from_args() (in module wrf), 190

219
wrf-python Documentation, Release 1.2.0

from_var (class in wrf), 190 get_w_destag() (in module wrf.g_wind), 176

get_xtimes() (in module wrf.g_times), 169
G getproj() (in module wrf), 132
geo_bounds() (in module wrf), 82 getvar() (in module wrf), 61
GeoBounds (class in wrf), 136
get_2dcape() (in module wrf.g_cape), 146 H
get_3dcape() (in module wrf.g_cape), 147 has_time_coord() (in module wrf), 128
get_avo() (in module wrf.g_vorticity), 173
get_basemap() (in module wrf), 84 I
get_cache_size() (in module wrf), 126 interp1d() (in module wrf), 90
get_cached_item() (in module wrf), 133 interp2dxy() (in module wrf), 91
get_cartopy() (in module wrf), 83 interplevel() (in module wrf), 68
get_cloudfrac() (in module wrf.g_cloudfrac), 148 interpline() (in module wrf), 70
get_coord_pairs() (in module wrf), 127 interpz3d() (in module wrf), 92
get_ctt() (in module wrf.g_ctt), 149 is_coordvar() (in module wrf), 128
get_dbz() (in module wrf.g_dbz), 150 is_mapping() (in module wrf), 128
get_destag_wspd_wdir() (in module wrf.g_wind), 177 is_moving_domain() (in module wrf), 129
get_destag_wspd_wdir10() (in module wrf.g_wind), 178 is_multi_file() (in module wrf), 127
get_dp() (in module wrf.g_dewpoint), 152 is_multi_time_req() (in module wrf), 127
get_dp_2m() (in module wrf.g_dewpoint), 153 is_staggered() (in module wrf), 130
get_eth() (in module wrf.g_temp), 164 is_standard_wrf_var() (in module wrf), 130
get_geopt() (in module wrf.g_geoht), 154 is_time_coord_var() (in module wrf), 127
get_height() (in module wrf.g_geoht), 154 iter_left_indexes() (in module wrf), 131
get_id() (in module wrf), 132 IterWrapper (class in wrf), 192
get_iterable() (in module wrf), 129
get_left_indexes() (in module wrf), 130 L
get_max_dbz() (in module wrf.g_dbz), 151 LambertConformal (class in wrf), 141
get_omega() (in module wrf.g_omega), 157 lat (wrf.CoordPair attribute), 135
get_pressure() (in module wrf.g_pressure), 158 LatLon (class in wrf), 144
get_pressure_hpa() (in module wrf.g_pressure), 158 latlon_coords() (in module wrf), 83
get_proj_params() (in module wrf), 131 latlon_coordvars() (in module wrf), 128
get_pvo() (in module wrf.g_vorticity), 174 latlon_str() (wrf.CoordPair method), 135
get_pw() (in module wrf.g_pw), 159 left_iteration() (in module wrf.decorators), 179
get_pyngl() (in module wrf), 85 ll_points() (in module wrf), 133
get_rh() (in module wrf.g_rh), 160 ll_to_xy() (in module wrf), 73
get_rh_2m() (in module wrf.g_rh), 161 ll_to_xy_proj() (in module wrf), 75
get_right_slices() (in module wrf), 131 lon (wrf.CoordPair attribute), 135
get_slp() (in module wrf.g_slp), 161
get_srh() (in module wrf.g_helicity), 155 M
get_tc() (in module wrf.g_temp), 167
map_proj (wrf.WrfProj attribute), 137
get_temp() (in module wrf.g_temp), 163
Mercator (class in wrf), 142
get_terrain() (in module wrf.g_terrain), 167
moad_cen_lat (wrf.WrfProj attribute), 137
get_theta() (in module wrf.g_temp), 162
get_times() (in module wrf.g_times), 168
get_tk() (in module wrf.g_temp), 166
N
get_tv() (in module wrf.g_temp), 164 npbytes_to_str() (in module wrf), 129
get_tw() (in module wrf.g_temp), 165 NullProjection (class in wrf), 141
get_u_destag() (in module wrf.g_wind), 174
get_uh() (in module wrf.g_helicity), 156 O
get_uvmet() (in module wrf.g_uvmet), 169 omega() (in module wrf), 111
get_uvmet10() (in module wrf.g_uvmet), 170 omp_destroy_lock() (in module wrf), 122
get_uvmet10_wspd_wdir() (in module wrf.g_uvmet), 172 omp_destroy_nest_lock() (in module wrf), 122
get_uvmet_wspd_wdir() (in module wrf.g_uvmet), 171 omp_enabled() (in module wrf), 113
get_v_destag() (in module wrf.g_wind), 175 omp_get_active_level() (in module wrf), 121

220 Index
wrf-python Documentation, Release 1.2.0

omp_get_ancestor_thread_num() (in module wrf), 120 set_cloudfrac_alg_metadata() (in module

omp_get_dynamic() (in module wrf), 115 wrf.metadecorators), 188
omp_get_level() (in module wrf), 120 set_cloudfrac_metadata() (in module
omp_get_max_active_levels() (in module wrf), 120 wrf.metadecorators), 185
omp_get_max_threads() (in module wrf), 114 set_destag_metadata() (in module wrf.metadecorators),
omp_get_nested() (in module wrf), 119 188
omp_get_num_procs() (in module wrf), 114 set_height_metadata() (in module wrf.metadecorators),
omp_get_num_threads() (in module wrf), 117 185
omp_get_schedule() (in module wrf), 116 set_interp_metadata() (in module wrf.metadecorators),
omp_get_team_size() (in module wrf), 121 186
omp_get_thread_limit() (in module wrf), 116 set_latlon_metadata() (in module wrf.metadecorators),
omp_get_thread_num() (in module wrf), 118 185
omp_get_wtick() (in module wrf), 116 set_uvmet_alg_metadata() (in module
omp_get_wtime() (in module wrf), 116 wrf.metadecorators), 187
omp_in_final() (in module wrf), 121 set_wind_metadata() (in module wrf.metadecorators),
omp_in_parallel() (in module wrf), 118 184
omp_init_lock() (in module wrf), 121 slp() (in module wrf), 93
omp_init_nest_lock() (in module wrf), 122 smooth2d() (in module wrf), 97
omp_set_dynamic() (in module wrf), 114 srhel() (in module wrf), 105
omp_set_lock() (in module wrf), 122 stand_lon (wrf.WrfProj attribute), 137
omp_set_max_active_levels() (in module wrf), 119
omp_set_nest_lock() (in module wrf), 122 T
omp_set_nested() (in module wrf), 119 td() (in module wrf), 94
omp_set_num_threads() (in module wrf), 114 tk() (in module wrf), 94
omp_set_schedule() (in module wrf), 115 to_np() (in module wrf), 78
omp_test_lock() (in module wrf), 123 top_right (wrf.GeoBounds attribute), 136
omp_test_nest_lock() (in module wrf), 123 truelat1 (wrf.WrfProj attribute), 137
omp_unset_lock() (in module wrf), 123 truelat2 (wrf.WrfProj attribute), 137
omp_unset_nest_lock() (in module wrf), 123 tvirtual() (in module wrf), 110

P U
pairs (wrf.combine_dims attribute), 189 udhel() (in module wrf), 106
pairs_to_latlon() (in module wrf), 134 uvmet() (in module wrf), 96
PolarStereographic (class in wrf), 143 uvmet_left_iter() (in module wrf.specialdec), 182
pole_lat (wrf.WrfProj attribute), 137
pole_lon (wrf.WrfProj attribute), 137 V
proj4() (wrf.WrfProj method), 140 varname (wrf.from_var attribute), 190
psafilepath() (in module wrf), 132 varnames (wrf.either attribute), 189
pvo() (in module wrf), 108 vertcross() (in module wrf), 68
pw() (in module wrf), 112 vinterp() (in module wrf), 72
pyngl() (wrf.WrfProj method), 140
pyngl_enabled() (in module wrf), 125 W
R wetbulb() (in module wrf), 110
WrfProj (class in wrf), 137
rh() (in module wrf), 95
RotatedLatLon (class in wrf), 144 X
S x (wrf.CoordPair attribute), 135
xarray_enabled() (in module wrf), 124
set_alg_metadata() (in module wrf.metadecorators), 186 xy() (in module wrf), 89
set_cache_size() (in module wrf), 126 xy_str() (wrf.CoordPair method), 136
set_cape_alg_metadata() (in module wrf.metadecorators), xy_to_ll() (in module wrf), 74
187 xy_to_ll_proj() (in module wrf), 76
set_cape_metadata() (in module wrf.metadecorators), 185

Index 221
wrf-python Documentation, Release 1.2.0

Y
y (wrf.CoordPair attribute), 135

222 Index

Meteorology Today: An Introduction To Weather, Climate, and The Environment 13Th Edition C. Donald Ahrens - Ebook PDF
100% (8)
Meteorology Today: An Introduction To Weather, Climate, and The Environment 13Th Edition C. Donald Ahrens - Ebook PDF
41 pages
Manual Jungheinrich DFG 540-550
100% (2)
Manual Jungheinrich DFG 540-550
171 pages
Verification of WRF Simulation PDF
No ratings yet
Verification of WRF Simulation PDF
8 pages
WMO Standard
No ratings yet
WMO Standard
118 pages
WRF Tutorial
No ratings yet
WRF Tutorial
84 pages
Asl410 WRF
No ratings yet
Asl410 WRF
47 pages
ARWUsersGuideV3 WRF Model
No ratings yet
ARWUsersGuideV3 WRF Model
411 pages
Description of The WRF-1d PBL Model
No ratings yet
Description of The WRF-1d PBL Model
11 pages
Numerical Weather Prediction (NWP)
No ratings yet
Numerical Weather Prediction (NWP)
20 pages
WRFUsersGuide PDF
No ratings yet
WRFUsersGuide PDF
456 pages
Radiation Parameterization Intro PDF
No ratings yet
Radiation Parameterization Intro PDF
79 pages
Get Geophysical Convection Dynamics Jun Ichi-Yano (Author) free all chapters
100% (3)
Get Geophysical Convection Dynamics Jun Ichi-Yano (Author) free all chapters
55 pages
Also: Hurricanes, Typhoons
No ratings yet
Also: Hurricanes, Typhoons
20 pages
WRF Modeling System Overview
No ratings yet
WRF Modeling System Overview
37 pages
Introductory Remarks: Introduction To Numerical Weather Prediction
No ratings yet
Introductory Remarks: Introduction To Numerical Weather Prediction
16 pages
Weather Forecasting Satellites
No ratings yet
Weather Forecasting Satellites
29 pages
WRF Installation
No ratings yet
WRF Installation
4 pages
Statistical Downscaling Portal
No ratings yet
Statistical Downscaling Portal
20 pages
Footprints in Micrometeorology and Ecology (PDFDrive)
100% (1)
Footprints in Micrometeorology and Ecology (PDFDrive)
254 pages
Thunderstorms and Tornadoes - Abogado, Loreen Jane
No ratings yet
Thunderstorms and Tornadoes - Abogado, Loreen Jane
45 pages
WRF Nesting
No ratings yet
WRF Nesting
46 pages
Mass 369104ngs
No ratings yet
Mass 369104ngs
5 pages
Cloud Physics Review - Includes Cloud Physics and Cloud Dynamics
No ratings yet
Cloud Physics Review - Includes Cloud Physics and Cloud Dynamics
5 pages
Sbdart User Manual Iisc
100% (1)
Sbdart User Manual Iisc
6 pages
Meteorology10 WeatherintheTropics
No ratings yet
Meteorology10 WeatherintheTropics
49 pages
Synoptic Meteorology
100% (1)
Synoptic Meteorology
40 pages
Synoptic Meteorology-Lecture12
No ratings yet
Synoptic Meteorology-Lecture12
29 pages
Applications of GIS For Disaster EWS by Jacob Opadeyi
No ratings yet
Applications of GIS For Disaster EWS by Jacob Opadeyi
61 pages
WRF Tutorial
No ratings yet
WRF Tutorial
49 pages
Lecture04 Atmosphere
No ratings yet
Lecture04 Atmosphere
29 pages
Mesoscale Convective System
No ratings yet
Mesoscale Convective System
7 pages
Climate Change Ques
No ratings yet
Climate Change Ques
26 pages
10.1175 BAMS ExplainingExtremeEvents2014.2
No ratings yet
10.1175 BAMS ExplainingExtremeEvents2014.2
180 pages
Weather Forecasting and Prediction
100% (1)
Weather Forecasting and Prediction
24 pages
Installation of Weather Research and Forecasting (WRF) Model On Linux Operating System
No ratings yet
Installation of Weather Research and Forecasting (WRF) Model On Linux Operating System
7 pages
04 Fronts and Mid Latitude Cyclones
No ratings yet
04 Fronts and Mid Latitude Cyclones
28 pages
Synop
No ratings yet
Synop
14 pages
Tutorial WRF
No ratings yet
Tutorial WRF
20 pages
Meteorology CYCLONES BSC
No ratings yet
Meteorology CYCLONES BSC
57 pages
Measurement and Analysis of Atmospheric Stability
No ratings yet
Measurement and Analysis of Atmospheric Stability
16 pages
Getting Your Hands-On Climate Data - Visualize Climate Data With Python
No ratings yet
Getting Your Hands-On Climate Data - Visualize Climate Data With Python
20 pages
Air Masses and Fronts
No ratings yet
Air Masses and Fronts
17 pages
COAWST User Manual
No ratings yet
COAWST User Manual
87 pages
Pres Leeway
No ratings yet
Pres Leeway
32 pages
Ocean Circulation
No ratings yet
Ocean Circulation
46 pages
NCL Functions and Procedures Reference Cards
No ratings yet
NCL Functions and Procedures Reference Cards
13 pages
Tutorial Import Excel Ke QGis
No ratings yet
Tutorial Import Excel Ke QGis
7 pages
Intro To Remote Sensing Test Questions
No ratings yet
Intro To Remote Sensing Test Questions
3 pages
BeachHazards RipCurrents0712
No ratings yet
BeachHazards RipCurrents0712
4 pages
Burned Area Mapping With Sentinel-2 (Snap)
No ratings yet
Burned Area Mapping With Sentinel-2 (Snap)
33 pages
Plot Synop and Ships
No ratings yet
Plot Synop and Ships
6 pages
Lecture 2 Wind Characteristics1 1 PDF
No ratings yet
Lecture 2 Wind Characteristics1 1 PDF
31 pages
GMT For Windows Installing
100% (1)
GMT For Windows Installing
28 pages
(Original PDF) Meteorology Today 11th Edition by C. Donald Ahrens 2024 Scribd Download
100% (8)
(Original PDF) Meteorology Today 11th Edition by C. Donald Ahrens 2024 Scribd Download
41 pages
Weather Forcast System: Case Study By-Gajendra Singh 12 F
100% (1)
Weather Forcast System: Case Study By-Gajendra Singh 12 F
14 pages
Ah Made Tal Experimental Agro Meteorology
No ratings yet
Ah Made Tal Experimental Agro Meteorology
159 pages
Prospects of Microwave Remote Sensing For Snow Hydrology: Helmut Rott
No ratings yet
Prospects of Microwave Remote Sensing For Snow Hydrology: Helmut Rott
10 pages
Analytic Surfaces
No ratings yet
Analytic Surfaces
109 pages
Weather Radar Polarimetry 1st Edition Guifu Zhang download pdf
100% (13)
Weather Radar Polarimetry 1st Edition Guifu Zhang download pdf
60 pages
Satellite Meteorology,
From Everand
Satellite Meteorology,
R. R. Kelkar
5/5 (1)
Elementary Positional Astronomy Supported by PC Software Programs
From Everand
Elementary Positional Astronomy Supported by PC Software Programs
ArturoandRaffaele Chiesa
5/5 (1)
Learning Style Vak
No ratings yet
Learning Style Vak
3 pages
Notes Technical Writing
No ratings yet
Notes Technical Writing
4 pages
Log 240922
No ratings yet
Log 240922
98 pages
Diabetes Mellitus NCP
No ratings yet
Diabetes Mellitus NCP
5 pages
Financial Analysis: P&G vs. Colgate-Palmolive
No ratings yet
Financial Analysis: P&G vs. Colgate-Palmolive
24 pages
Linguistik Assigment by Sonia Fitri-1
No ratings yet
Linguistik Assigment by Sonia Fitri-1
6 pages
21st Century Skills PDF
100% (1)
21st Century Skills PDF
15 pages
TheLevelofSpendingHabitsamongAccountancyBusinessandManagementStudentsofTacurongNationalHighSchoolBasisforProgramIntervention PDF
No ratings yet
TheLevelofSpendingHabitsamongAccountancyBusinessandManagementStudentsofTacurongNationalHighSchoolBasisforProgramIntervention PDF
48 pages
txl212 Slides Part1
No ratings yet
txl212 Slides Part1
71 pages
Instructional Design Project
No ratings yet
Instructional Design Project
13 pages
BMPP-KL2-01-SPC-ME-002-Rev. 0-Specification For Engine
No ratings yet
BMPP-KL2-01-SPC-ME-002-Rev. 0-Specification For Engine
13 pages
Efa - Faq
No ratings yet
Efa - Faq
10 pages
AUTOSAR Runtime Environment and Virtual Function Bus: Nico Naumann
No ratings yet
AUTOSAR Runtime Environment and Virtual Function Bus: Nico Naumann
19 pages
Guidebook For International Students (2020 Spring Semester Freshmen) 2
No ratings yet
Guidebook For International Students (2020 Spring Semester Freshmen) 2
40 pages
George G. McBride - RSA03-The Art of Penetration Testing
No ratings yet
George G. McBride - RSA03-The Art of Penetration Testing
59 pages
Journal of Experimental Zoology Part A Comparative Experimental Biology - 2004 - Steinberg - Townes and Holtfreter 1955
No ratings yet
Journal of Experimental Zoology Part A Comparative Experimental Biology - 2004 - Steinberg - Townes and Holtfreter 1955
6 pages
Chemistry Viva-Voce PDF
No ratings yet
Chemistry Viva-Voce PDF
11 pages
Communication For Various Purposes Quiz
100% (1)
Communication For Various Purposes Quiz
2 pages
Syllabus RRB JE
No ratings yet
Syllabus RRB JE
3 pages
AdditionalAsstProb Lecture02 ProjMgmt
No ratings yet
AdditionalAsstProb Lecture02 ProjMgmt
1 page
Wells HSE MS Manual
No ratings yet
Wells HSE MS Manual
47 pages
APA Style Reerence Citations
No ratings yet
APA Style Reerence Citations
4 pages
Entrepreneurs Are Great Worksheets
No ratings yet
Entrepreneurs Are Great Worksheets
9 pages
United States Patent (10) Patent No.: US 7.931,757 B2
No ratings yet
United States Patent (10) Patent No.: US 7.931,757 B2
14 pages
Envir Otect Power Cable Install Manual
No ratings yet
Envir Otect Power Cable Install Manual
49 pages
KB Rebar Brochure
100% (1)
KB Rebar Brochure
4 pages
Welding QC Q&A
No ratings yet
Welding QC Q&A
6 pages
The Anatomy of A Mutual Fund
No ratings yet
The Anatomy of A Mutual Fund
6 pages
Alloying Elements Excel
No ratings yet
Alloying Elements Excel
18 pages