AI/ML methods applied to VLF/LF measurements

In this study we investigate the application of Artificial Intelligence (AI) Machine Learning (ML) methods – so called Transformer Architectures – to very low frequency/low frequency (VLF/LF) data sets. The research is based on electric field measurements from VLF/LF receivers of the INFREP network. The scientific objective is to characterize physical phenomena, such as variations in the electromagnetic and particle environment (e.g., solar x-ray flares), changes in ionospheric plasma parameters, and impact on the ionosphere from extraterrestrial events (e.g., gamma-ray bursts). Standard data processing algorithms in the time- and frequency domain are used to cross-check the AI/ML results. We give an overview of the status of the project, show preliminary results and discuss pros and cons of different AI/ML approaches applied to VLF/LF data.