Authors: Cooper Downs (Predictive Science Inc), Jon A. Linker (Predictive Science Inc), Ronald M. Caplan (Predictive Science Inc), and the Solo/PHI team (MPS)

The key observational input for data constrained models of the solar corona is the boundary condition for the normal component of magnetic field, Br, at the coronal base. Frequently, Br is derived from LOS magnetograms or maps, as these are often the most easily (and sometimes only) accessible data. Typically, Br is derived from the LOS component of the field, Blos, under the assumption that Blos is predominantly radial in the photosphere where it is measured (e.g. Wang & Sheeley 1992). This geometric approximation often performs well in the weaker field regions of the Sun, but it can be a poor approximation in sunspots, where strongly non-radial fields are clearly present. Another option, used in the earliest potential field models (Altshuler & Newkirk 1969) is to use the machinery of 3D potential field extrapolations to derive the Br distribution of the potential field whose LOS projection will match Blos. As illustrated in Leka et al. (2017) this may have important consequences for the strength and structure of the inferred Br in solar active regions (ARs), and is particularly relevant for artifacts that appear in strong field umbral and penumbral regions observed off of the Sun-Earth line. Leka et al. used a spherical harmonic approach. Here we describe our recent efforts to develop a pipeline for computing the LOS matching potential field using our high-performance finite-difference potential field solver POT3D. Using a simple iterative method we are able to overcome some of the limitations of earlier approaches while adding the flexibility to localize the calculation, e.g., to use the geometric approximation to derive Br in weaker field regions, while imposing the Blos boundary condition in stronger field regions. We illustrate the practical relevance of these considerations by applying the technique to the case that motivated this work, the backside CME event of September 5, 2022. In this case the only available magnetic field measurements of the rapidly evolving source region are LOS magnetograms from the SolO/PHI/FDT instrument, and this region was ~40 degrees away from disk center as seen by SolO at the time of the eruption. By computing the Br boundary condition in various ways, we illustrate how each technique brings along its own set of issues and how these may be partially ameliorated using a hybrid approach with localization. We then explore how solutions for the global coronal field are impacted by these choices, including non-negligible changes to the footprint of open flux and the S-WEB structure. While there is clearly no substitute for vector magnetic field information (when available), we also discuss future applications where this practical technique may be relevant.