Geo-Application Development: Collocated EnMAP Sentinel-5P

Summary #

This project supports work conducted for my master’s thesis. The thesis investigates the use of neural networks trained on Sentinel-5P/TROPOMI NO₂ slant columns to retrieve atmospheric NO₂ concentrations from the EnMAP hyperspectral imager. A necessary step to achieve this is the identification of collocated EnMAP and TROPOMI acquisitions that are used to train a neural network model. The temporal offset between acquisitions should be minimized so that both instruments capture almost the same conditions. The code developed in this project provides a Python workflow to

1. identify collocated EnMAP and TROPOMI acquisitions,
2. identify the temporal offsets between acquisitions,
3. export EnMAP tile geometries and offsets to the temporally closest TROPOMI acquisition as a GeoPackage file,
4. and visualize the results usings PyQGIS.

A detailed overview of the project including all code, resources, and documentation can be found on GitHub.