C22

Presentation Title: Flare Forecasting and SolarMonitor 2
Author(s): Shane Maloney, Sophie Murray, Peter Gallagher, Paul Wright, Anna Massone, Michele Piana, Tamas Kiss, Gabriele Pierantoni, Gabor, Robert Lovas, Attila Farkas

Abstract:

The Sun is an enigmatic star that produces the most powerful explosive events in our solar system – solar flares and coronal mass ejections. Studying these phenomena can provide a unique opportunity to develop a deeper understanding of fundamental processes on the Sun, and critically, to better forecast space weather. The Active Region Classification and Flare Forecasting (ARCAFF) project will develop a beyond state-of-the-art flare forecasting system utilising end-to-end deep learning (DL) models to significantly improve upon traditional flare forecasting capabilities.

The five key objectives of ARCAFF are:
1. Active region classifications using magnetogram cutouts
2. Active region localisation and classification using full disk magnetograms
3. Point-in-time flare prediction using full disk magnetograms
4. Point-in-time flare prediction using full disk multimodal observations
5. Time series flare prediction based on time series of full disk multimodal observation

The huge amount of space-based solar observations are an ideal candidate for this type of analysis, given DL effectiveness in modelling complex relationships. DL has already been successfully developed and deployed in weather forecasting, financial services, and health care domains, but has not been fully exploited in the solar physics domain. The forecasts will be benchmarked against current systems using international community standards.

Benchmarking and intercomparison is a key challenge in solar flare forecasting and in particular machine learning based solutions. As part of the ARCAFF project and the ongoing SolarMonitor2 upgrade. SolarMonitor2 will run and store the results from the ARCAFF and other flare forecast models. Based on the results standard metrics will be computed and displayed - in particular on unseen data - allowing for fair and unbiased comparisons.