Title: To E or not to E: Numerical Nuances of Global Coronal Models

Abstract: In the recent years, global coronal models have experienced an ongoing increase in popularity as tools for forecasting solar weather. Within the domain of up to 21.5Rsun, magnetohydrodynamics (MHD) is used to resolve the coronal structure using magnetograms as inputs at the solar surface. Ideally, these computations would be repeated with every update of the solar magnetogram so that they could be used in the ESA Modelling and Data Analysis Working Group (MADAWG) magnetic connectivity tool (this http URL). Thus, it is crucial that these results are both accurate and efficient. While much work has been published showing the results of these models in comparison with observations, not many of it discusses the intricate numerical adjustments required to achieve these results. These range from details of boundary condition formulations to adjustments as large as enforcing parallelism between the magnetic field and velocity. By omitting the electric field in ideal-MHD, the description of the physics can be insufficient and may lead to excessive diffusion and incorrect profiles. We formulate inner boundary conditions which, along with other techniques, reduce artificial electric field generation. Moreover, we investigate how different outer boundary condition formulations and grid design affect the results and convergence, with special focus on the density and the radial component of the B-field. The significant improvement in accuracy of real magnetic map-driven simulations is illustrated for an example of the 2008 eclipse.