Pytroll: Easily Accessing and Exploiting Weather Satellite Data ( Speaker: Martin Raspaud SMHI)

The Pytroll project is comprised of almost 20 free and open source python modules to read, analyse, process and write weather satellite data. Users and developers from all around the world are contributing to make Pytroll the go-to python framework for weather satellite imager data. At the same time, Pytroll is written with modularity and genericity in mind, allowing it to be used in other contexts within earth science.

In this seminar, we will first present the Pytroll modules, covering data reading, combining, remapping, decorating among other things.

Then we will give real-world examples (read actual python code examples) on how to perform different weather satellite related tasks, using both polar orbiting and geostationary imager data to work on, and show how the unified framework simplifies handling data from different satellites.

Then, we will present how to batch-process data, showing how pytroll is used (in combination with other python modules) for stable 24/7 operations at the Swedish Meteorological and Hydrological Institute, using message queues for low-latency real-time processing of weather satellite data.

Finally, we will discuss some tools and methods we use when developing Pytroll modules, like continuous integration, test driven development, code quality monitoring, code versioning and communication means.

More information at www.pytroll.org