Welcome to trollBUFR’s documentation!


The Python package trollbufr provides an API and some command-line scripts to read/decode and write/encode BUFR messages.

Where comes the troll from? – Well, this package lives in the context of PyTroll, an Open-Source-Software collaboration of several national meteorological services and universities.

About BUFR

BUFR stands for “Binary Universal Format for data Representation”. It is a binary message format, developed as a “table-driven code form” (TDCF) by members of World Meteorological Organisation (WMO).

It’s main use is the meteorological data exchange and storage. It is used in other fields of geo-science as well, e.g. oceanographics, and satellite-derived products.

Find further information and detailed description at http://www.wmo.int/pages/prog/www/WMOCodes.html


With version 0.6.0 a feature for encoding a JSON formatted file into binary BUFR is added.

The command-line options are changed:

  • print information and descriptor sections: -d became -m.
  • decode BUFR: -r became -d.

New is -j to write any output in a JSON format.


There are still things to do:

  • Create working unit-tests.
  • Implement the remaining obscure operators.

So, get involved at PyTroll or GitHub!


trollBUFR is licensed under GNU LGPL-3.

Indices and tables