Workshop 2013

pict1 pict2
After a two day open workshop we were 7 pytrollers from Finland, Iceland and Sweden staying till the end of week working together on some pressing issues. See below for a summary of achievements. If you would like to contribute actively with the pytroll development, please let us know at the mailing list ( We plan to have one or two pytroll weeks (usually 3-4 days of dedicated programming) next year. Usually we will identifiy a few specific topics that we think needs special attention.

Summary achievements

  • Extending the user community: Several new users have become more familiar with Pytroll and started contributing.
  • Testing: A number of bugs and user inconveniences were identified, and some have been solved already.
  • Enhancements to Pyresample:
    • Now Pyresample allows to attach a weight to the gaussian reprojection method. This is convenient when e.g. gridding several swath products into a level 2.5/3 product (Climate applications).
  • MIPP enhancements and user documentation: MIPP allows XRIT decompression on the fly, and MIPP documentation slightly improved.
  • Three new projects initiated:
    • Pydecorate to add logos, text, color bars and stuff to images
    • Trollimage - an enhancement of the module in mpop including some color enhancements. Will deprecate in mpop
    • Trollduction - A modular batch production framework for Pytroll
  • netCDF reader for SSM/I
  • Trollcast testing, for data exchange between SMHI and FMI - resolution is pending (time outs)
  • Designing and developing the FMI Pytroll based polar production system. Probably resulting in a general concept that can be useful to other users
  • Looking at how to enable web based batch production monitoring with Pytroll. Could e.g. be used for an easy and quick access to Pytroll products for in-house R&D
  • Initiated an overhaul and check of the EUMETSAT recipe RGBs in Pytroll and how they compare and deviate with the official ones. Done in collaboration with the Romanian Met Service.
  • Colorizing Pytroll images (using the new trollimage component) - ‘’sandwich’’ product. Color enhanced imagery is commonly used in forecast offices, e.g. IR imagery with cold temperatures enhanced using a color palette.