Skip to content

observingClouds/eurec4a_snd

Repository files navigation

eurec4a_snd

Software Data

During the EUREC4A/ATOMIC campaign many radiosondes were launched from different platforms and partners. This collection of scripts has been used to convert, visualize and interpolate the data in a common way.

This repository has been created before the start of the campaign and evolved further in the postprocessing phase. It therefore contains scripts to

  • convert and visualize BUFR and Vaisala MW41 rawdata (.mwx)
  • interpolate sounding data
  • correct and quality control soundings

A generalized version of this repository has evolved at https://github.com/observingClouds/pysonde.

While most of the scripts can be applied to every sounding that exists in the BUFR or mwx format, some of the postprocessing scripts are specific to the EUREC4A campaign and can only serve as a guidance for other studies.

What's new

Changes to the software and produced data can be found here

Installation

The package can be installed with conda (Miniconda or Anaconda):

The best option is to create a new environment to not run into dependency problems with pre-installed package:

conda create --name field_campaign

Activate the new environment with

conda activate field_campaign

or

source activate field_campaign

The actual package can than be installed with

conda install -c observingclouds -c conda-forge eurec4a_snd

First execution

To convert BUFR files, simply run sounding_converter -i your_sounding.bfr after the installation. During the first execution you will be asked to give meta-information which will be included in the output files.

These include e.g. contact_person and platform name which are probably helpful after the exchange of these files with other scientists.

It is also possible to copy the config/meta_information_template.ini to one's favorite folder and change the files content accordingly. If the file is not renamed meta_information.ini and is in the home directory, the option -c has to be used during the call of sounding_converter otherwise it will not recognize the configuration.

(macOS or linux)

sounding_converter -i your_sounding.bfr -c /your/path/to/meta_information.ini

(windows)

sounding_converter.exe -i your_sounding.bfr -c /your/path/to/meta_information.ini

Update

Please check for updates at the beginning of the campaign by running

conda install -c observingclouds -c conda-forge eurec4a_snd

Change meta-information

The meta-information which is used in the netCDF files and in order to create a reasonable filename, can always be changed by either

  • remove the file meta_information.ini in $HOME and run sounding_converter again, as if it would be your first execution
  • edit the file ~/meta_information.ini directly

Visualization

Panoply

The converted dat files are netCDF files which conform to the CF-Conventions as far as possible and make use of the discrete sampling geometry. The sounding data can therefore easily be drawn as trajectories without extra efforts. One example is the software Panoply

Trajectory visualization with panoply

Simple Plotting

The package also includes a few plotting routines that can be called with e.g.

(macOS or linux)

sounding_visualize -i converted/file/sounding.nc

(windows)

sounding_visualize.exe -i converted/file/sounding.nc

SkewT Plotting

A skewT diagram can be created with

(macOS or linux)

sounding_skewT -i converted/file/sounding.nc

(windows)

sounding_skewT.exe -i converted/file/sounding.nc

Further examples on how to create a skewT diagram can be found in eurec4a_snd/examples/visualizing. Have a look at Skew-T examples

Example

Examples of the input .bfr files and the converted .nc file can be found in examples/data.

Interpolation

To concatenate the soundings and for an easier analysis, the converted soundings (output of sounding_converter) can be interpolated on a common altitude grid by

sounding_interpolate -i /path/to/converted/soundings*.nc

Trouble shooting

ImportError: DLL load failed: The specific module could not be found
Windows users might get the above error message when trying to visualize the soundings. The error is caused in pillow. Unfortunately there is not a very good solution yet, but the following might work for you:
conda remove --force pillow
pip install pillow

Commands `sounding_convert` and `sounding_visualize` cannot be found!
It seems something went wrong with the installation via `conda`. Although it is recommended to install this package via anaconda because it comes with the benefit that all dependencies should be resolved, you can also download this git repository and run within the `eurec4a_snd` folder:
python L1_rs41.py -i your_sounding.bfr
or for quicklooks
python make_quicklooks_rs41.py -n converted/file/sounding.nc

Slow internet connection: Download failed
In case of a slow internet connection, the command `conda install some_package` might fail due to connection timeout. In this case it might be a good option to download the failing package manually.

In this case the eurec4a_snd package for OSX is downloaded and installed manually:

wget -c https://anaconda.org/observingClouds/eurec4a_snd/v2.0.0/download/osx-64/eurec4a_snd-v2.0.0-py37_0.tar.bz2

The path needs to be adapted depending on the operating system and the version that should be downloaded. You may actually see the path you need to download in the error message of conda install eurec4a_snd.

The installation follows simply with

conda install -c observingClouds eurec4a_snd-v2.0.0-py37_0.tar.bz2
ECCODES_DEFINITION_PATH defined incorrectly
In case there appears an error, that the variable ECCODES_DEFINITION_PATH is not defined or not correctly set, the easiest solution is to insert the following two lines at the beginning of the script L1_bufr.py.
import os
os.environ["ECCODES_DEFINITION_PATH"] = r"C:\Users\...\envs\field_campaign\Library\share\eccodes\definitions"
The path needs to be adapted to your environment of course.