Solar flares are powerful explosions on the Sun that can knock out satellites and blackout cities. Predicting them is complicated by the tangled magnetic fields in sunspots. Scientists monitor the star by breaking its light into colors (spectroscopy) and measuring brightness (photometry). A new deep learning approach not only delivers a forecast but also reveals the logic behind its decision, like a chef explaining why the soup turned out so well.
The algorithm shows which magnetic field parameters were decisive. This turns the neural network's "black box" into a transparent guide: engineers can protect equipment faster, and physicists refine solar models. Interestingly, a single flare releases energy equivalent to billions of atomic bombs, but now we face it fully armed, knowing not just the threat but its source.
The energy of one powerful solar flare is comparable to the explosion of billions of atomic bombs or the eruption of tens of thousands of volcanoes.