Including the transition region and chromosphere in a global model for the solar atmosphere

In my ERC-AdG project ‘Open SESAME’ (project No 101141362), we aim to develop a time-evolving model for the entire solar atmosphere, including the chromosphere and transition region, based on a multifluid description. Currently, models are primarily steady, rely on a single-fluid description and include only the corona due to computational challenges. We plan to use time-evolving ion-neutral and ion-neutral-electron models. The multifluid approach will enable us to describe the intricate physics in the partially ionised chromosphere and quantify the transfer of momentum and energy between the atmospheric layers. The questions of where the solar wind originates and how solar flares and coronal mass ejections are driven have fundamental scientific importance and substantial socio-economic impact.

This goal is now achievable by combining our implicit numerical solver with a high-order flux reconstruction (FR) method. The implicit solver enables larger time steps, avoiding the numerical instabilities that lead to strict time-step limitations in explicit schemes. The high-order FR method enables high-fidelity simulations on very coarse grids, even in zones of high gradients. We will introduce three critical innovations. First, we will combine high-order FR with physics-based r-adaptive (moving) unstructured grids, redistributing grid points toward regions with high gradients while preserving the HPC cluster's load-balancing. Second, we will implement CPU-GPU algorithms for the new heterogeneous supercomputers advanced by HPC-Europa. Third, we will implement AI-generated magnetograms to enable the model to respond to the time-varying photospheric magnetic field, which is crucial for understanding key properties of the solar plasma and processes.

Thus, we will develop a first-in-its-kind high-order GPU-enabled 3D time-accurate solver for multifluid plasmas. If successful, we will implement the most advanced data-driven solar atmosphere model in an operational environment. The project commenced on September 1, 2024, and we have already obtained interesting results in time-dependent full-MHD corona modelling, inclusion of the TR, AI-generated magnetograms (for the far side of the Sun), and high-order flux reconstruction simulations.