Presentation | A51B: AI-Driven Innovations in Earth and Atmospheric Sciences II Oral
Oral
[ONLINE] A51B-08: Retrieval of Cloud Properties for the Copernicus Atmospheric Missions Sentinel-4 and TROPOMI / Sentinel-5 Precursor using deep neural networks
Author(s): Fabian Romahn, German Aerospace Center (DLR) (First Author, Presenting Author) Diego Loyola, German Aerospace Center (DLR) Ronny Lutz, German Aerospace Center (DLR) Víctor Molina García, German Aerospace Center (DLR) Adrian Doicu, German Aerospace Center DLR Oberpfaffenhofen Athina Argyrouli, Technical University of Munich (TUM)
In order to cope with the vast amounts of data and requirements for timeliness, new artificial intelligence (AI) methods, in particular neural networks (NNs), are used to retrieve the cloud properties for the Copernicus atmospheric satellite missions.
We present two main methods of how NNs can be used in this context, their respective properties and their potential for an operational use.
