Applications of Atmospheric Composition Data: Open Source Training Materials by EUMETSAT

This presentation provides an overview of the open source training materials produced by the EUMETSAT Atmospheric Composition training team. The training materials covered in this presentation include: (1) LTPy - the Learning Tool for Python on Atmospheric Composition Data, (2) FANGS - Fire Applications with Next-Generation Satellites, (3) Dust Aerosol Detection, Monitoring and Forecasting, and (4) a self-paced training course on Identifying and Quantifying Dust using Satellite Data. 

The first three sets of training materials were developed using Jupyter notebooks, which allow for a high-level of interactive learning, as it makes code, instructions and visualisations available in the same location. Executable notebooks are available on a dedicated Jupyterhub-based course platform which has the required programming environment and data already preinstalled. In addition, an accompanying Jupyter Book is also available for two of the training modules. The final self-paced training course consists of a series of mini-modules on the Moodle platform.

The Learning tool for Python on Atmospheric Composition Data is a Python-based training course on Atmospheric Composition Data. The training course covers notebooks on data access, handling and processing, visualisation, case studies and exercises. LTPy features data from six different satellites, including the Copernicus satellites Sentinel-3 and Sentinel-5 as well as the polar-orbiting meteorological satellite series, Metop, and five different model-based product types from the two Copernicus services on Atmosphere Monitoring (CAMS) and Emergency Management (CEMS). The course facilitates the uptake and use of atmospheric composition data and showcases possible application areas. 

FANGS - Fire Applications with Next-Generation Satellites features Python-based training material and application cases on fire detection and monitoring of the fire life-cycle. The training material makes use of proxy and simulated data, including data from precursor instruments of the Meteosat Third Generation (MTG) and EUMETSAT Polar System - Second Generation (EPS-SG) satellite missions. The training material consists of modular Jupyter notebook case studies on the 2020 wildfires in California, USA and the Mediterranean wildfires in 2021. In total, 24 notebooks were developed comprising five narrative notebooks and 19 workflow notebooks. 

The training course on ‘Dust Aerosol Detection, Monitoring and Forecasting’ provides a hands-on introduction to satellite-, ground- and model-based data used for dust monitoring and forecasting. This Python-based course is organised in three main chapters: (i) observations (satellite- and ground-based), (ii) forecast models and a (iii) practical case study. It features twelve different datasets derived from satellites, ground-based measurement networks and forecast models. The course material is developed in the form of well-described and modular Jupyter notebooks. In total, the course consists of 17 notebooks; 12 data workflows and five practical exercise notebooks.

This presentation finally introduces a self-paced training course on identifying and quantifying dust using satellite data. This course is targeted at two audiences, namely, forecasters and researchers. At the end of the self-paced course, learners would have gained the skills to either (1) visualise dust events using Level 1 and Level 2 satellite data or (2) plot and interpret a time series of dust aerosol optical depth (AOD).