Invited seminar | The active Sun: imaging and forecasting

Where and when

Rome, Italy – June 14th, 2022

speaker: Anna Maria Massone
affiliation: Dipartimento di Matematica, Università di Genova
date: June 14th, 2022
venue: Rome (Italy), Monte Porzio Catone – Istituto Nazionale di Astrofisica: Osservatorio Astronomico di Roma (OAR)

The active Sun: imaging and forecasting

abstract: The Sun is an enigmatic star that produces some of the most powerful explosive events in our solar system. Studying these eruptions can provide a unique opportunity to better understand fundamental processes on the Sun, and to understand their space weather impacts at Earth and throughout the solar system. In this talk I will discuss some inverse problems and machine learning-based approaches for the solution of image reconstruction, image processing and forecasting problems, offering great potential to investigate/learn the characteristics of the Sun-Earth system.
First, I will introduce an image reconstruction method based on Particle Swarm Optimization (PSO) for forward fitting visibilities measured by the Spectrometer/Telescope Imaging X-rays (STIX) on-board Solar Orbiter. Then, I will present an inverse diffraction procedure to recover information in the significant amount of images provided by the Atmospheric Imaging Assembly in the Solar Dynamics Observatory (SDO/AIA) where saturation affects their most intense core, inhibiting their full exploitation. Finally, I will describe different approaches that can be followed to Space Weather forecasting. In this perspective by using data sets of almost 200 features extracted from the Helioseismic and Magnetic Imager (SDO/HMI) vector magnetograms, I will address the problem of flare forecasting so as the identification of the image properties that mostly impact the event prediction. I will also show results obtained following an innovative approach based on deep learning, whereby HMI images/videos can feed Convolutional Neural Networks (CNNs) that automatically extract peculiar features and perform probabilistic forecasting.


featured photo: Rawpixel Ltd, CC BY 2.0, via Wikimedia Commons

Last update 27 March 2023