Open Superior Efficient Solar Atmosphere Model Extension

The goal is to develop a time evolving model for the entire solar atmosphere, including the chromosphere and transition region, based on a multi-fluid description. At present, models are 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 multi-fluid approach will enable us to describe the intricate physics in the partially ionized chromosphere and quantize the transfer of momentum and energy between the atmospheric layers. The questions where the solar wind originates and solar flares and coronal mass ejections are driven have both fundamental scientific importance and substantial socio-economic impact. Indeed, the solar atmospheric model is the crucial missing link in the Sun-to-Earth model chain to predict the arrival and impact of CMEs at Earth.What makes this goal now possible is the combination of our implicit solver with a high-order flux-reconstruction (FR) method. The implicit solver avoids the numerical instabilities that lead to strict time step limitations on explicit schemes. The high-order FR method enables high-fidelity simulations on very coarse grids even in zones of high gradients. We will start from this new development and introduce three critical innovations. First, we will combine high-order FR with physics-based r-adaptive (moving) unstructured grids redistributing grid points to regions with high gradients. Second, we will implement CPU-GPU algorithms for the new heterogeneous supercomputers advanced by HPC-Europa. Third, we will implement AI generated magnetograms to make the model respond to the time-varying photospheric magnetic field which is crucial for understanding important properties.We will thus develop a first-in-its-kind high-order GPU-enabled 3D time-accurate solver for multi-fluid plasmas. If successful, we will have the most advanced solar atmosphere model implemented in an operational environment.