High-performance computing in Europe: Simulating solar magnetism

We work closely with the Partnership for Advanced Computing in Europe (PRACE), which provides leading scientists with access to the most powerful computers in Europe. Our new blog series will be showcasing this amazing work through stories we have written for their latest publication, the PRACE Digest 2021.

The blinding white disk in the sky we call the sun – the source of light and life on Earth – is a colossal, writhing ball of intensely hot plasma. Its surface spits with arcs of fire that extend for thousands of kilometers into space, caused by the fact that the sun is a giant magnetic star made of material that moves in concert with the laws of electromagnetism

The magnetism of the sun has a huge influence on our solar system, even causing visible effects on Earth in the form of auroras. It is produced by an enigmatic process called the solar dynamo which, to date, numerical models have only had moderate success in accurately describing. Overly simplistic models that do not take into account many of the processes known to occur in the sun have meant that a rough qualitative description has been the best that simulations have been able to offer.

A volume rendering of global convective dynamo simulations. Image credit Dr. Ameya Prabhu

Recently, more accurate observations of the sun have provided invaluable information to those working on simulations of the sun. The increasingly high-resolution images of the solar surface have revealed a seemingly unending depth of smaller and smaller magnetic structures, indicating that the solar plasma is not only intensively turbulent, but that this turbulence is likely an important constituent of solar magnetism.

Recently, top global dynamo modelling groups around the world have reached the point at which they can simultaneously capture the two main constituents of the solar dynamo, known as the large-scale and small-scale dynamos. Professor Maarit Käpylä of Aalto University has been leading a PRACE project with the objective of improving upon current numerical models to provide answers to the many questions and controversies that surround the solar dynamo. “The simplest upgrade that we were able to implement, thanks to the Tier-0 resources provided by PRACE, was to run the simulations at a higher resolution so that we could gain a better understanding of the turbulent flows and magnetic fields at play,” she says.

Aside from higher resolution, one of the major improvements made in the project was to remove over-simplistic models of heat conduction and instead allow it to evolve as a function of the density and temperature of the system. In the solar convection zone – the outermost layer of the solar interior – magnetism is generated in a cyclic fashion due to the motion of plasma. This is known as the large-scale dynamo. In the majority of previous models, the solar convection zone was assumed to be convective at all depths, but the new concept modelled by the team led by Käpylä removes this assumption.

As it turns out, tweaking the model to allow convection to evolve more freely had a big effect on the dynamics of the system. “With our improved model, we saw a layer at the bottom of the ‘convection’ zone that was not at all convective but still transported heat towards the surface – with entirely different properties than seen in previous simulations,” says Käpylä. “Interestingly, subsequent observational data published very recently seems to indicate that these layers do in fact exist. We had some doubts about our results when we carried out the analysis, but these observations have made us realise that we may have discovered something important in terms of magnetic field evolution in the sun.”

As well as the large-scale dynamo caused by the motion of plasma in the solar convection zone, there exists another dynamo instability called the small-scale dynamo which generates a non-cyclic, fluctuating component to the magnetic field of the sun. Until this project, the extent to which this small-scale dynamo occurred was not well known, but the increase in resolution and turbulence in these simulations enabled the researchers to confirm that this instability may well play an important role in the solar dynamo.

“Now that we have been able to locate and isolate the amplification of small-scale magnetic fields, we will be looking to study the influence of this effect on the overall solar dynamics,” says Käpylä. “It has been said that to truly understand the solar dynamo, this secondary instability must be accounted for, so we are quite excited that we will finally be able to investigate it properly.”

Professor Maarit Käpylä of Aalto University

Having participated in a number of PRACE projects in the past, Käpylä believes that this one has been her most successful to date. “When you carry out these extremely large simulations, you cannot store all of the data from them,” she says. “Instead, you have to harvest what you think is important and then move on. Consequently, in the aftermath you often have regrets and think about how you might have done things differently, but in this project, everything went just as planned.”

As the relentless march towards larger and more powerful computers and data centres continues, Käpylä believes that there is a need for more investment into optimising the codes that are used on them. “The world of high-performance computing is moving fast, and in the last few years we have already had to deal with our first major paradigm shift from using CPUs to GPUs,” she says. “With the advent of quantum computing next in line, we need to ensure that we are ready with algorithms that can adequately use these technologies.”

Read more about the pioneering science being done on Europe’s most powerful supercomputers

The HyperCOG project turns two
Previous Story
High-performance computing in Europe: Energetically excited states in heterogenous photocatalysts
Next Story
Want to read more like this? Subscribe to Projects Magazine today

Make your research count

Contact us now and let us help your research reach the right people