The massive explosive phenomena that are solar flares are caused by the accumulation of the twisting of the solar magnetic field (magnetic helicity), but it has previously been unclear what controls this twisting. Indeed, it was presumed that bundles of twist-free magnetic flux in the solar interior would not produce a solar flare.
The interior of the Sun remains invisible to telescopes, but activity can be probed using supercomputer simulations. The research team consisting of Assoc. Prof. Toriumi Shin (JAXA), Prof. Hotta Hideyuki, and Prof. Kusano Kanya (Nagoya University) used this technique to discover that thermal convection in the solar interior has a much greater impact than previously thought on the process that provides the magnetic helicity. The team performed large-scale numerical simulations using the supercomputers “Fugaku” (RIKEN) and “ATERUI II” (National Astronomical Observatory of Japan) to reproduce the emergence of bundles of magnetic flux (magnetic flux tubes) from the deep solar interior and their formation of sunspots. By artificially changing the strength of the twist of the flux tubes, the team investigated the differences in the process of injecting magnetic helicity into the solar corona as the magnetic flux emerges and produces sunspots.
The results of this revealed that even for the case with no magnetic twist, the sunspots rotated because the surrounding convection spins the magnetic flux, and magnetic helicity was injected into the solar corona. The amount of helicity supplied by the thermal convection reached a level that could cause lower-level solar flares. The results of this study indicate the possibility that not only the twist of the flux tubes but also the effect of thermal convection in spinning the magnetic flux may play a critical role in providing magnetic helicity and storing the energy required for solar flares, which is an achievement that updates the previous understanding. The results were published from Scientific Reports, a Nature portfolio journal.
Supernovae, exploding stars, play a critical role in the formation and evolution of galaxies. However, key aspects of them are notoriously difficult to simulate accurately in reasonably short amounts of time. For the first time, a team of researchers, including those from The University of Tokyo, apply deep learning to the problem of supernova simulation. Their approach can speed up the simulation of supernovae, and therefore of galaxy formation and evolution as well. These simulations include the evolution of the chemistry which led to life.