Doing an Ultrasound Scan of the Sun with AI | Monash Conversations on AI & Data Science

Monash Conversations on AI & Data Science series is presented by the Monash Data Futures Institute.
--------------------------------------------------

In the scenario of a strong solar coronal mass ejection, we would only receive a mere 30-minute warning before an immense solar event, which could have devastating consequences for our satellite technology and society. Regrettably, this limited time frame does not allow for the activation of emergency protocols or the implementation of practical measures.

To enhance our preparedness and minimise the potential impact, it is imperative to strive for significantly longer warning periods. It is essential that we lend our ears to the Sun's messages. In this session, Dr Alina Donea demonstrated the utilisation of solar images (generated by AI and a deep neural network, referring to conditional Generative Adversarial Network (sGAN)) to enhance predictions of solar activity from the side of the Sun that is not visible, leading to significant advancements in space weather forecasting.