Satabdwa Majumdar, Martin Reiss, Karin Muglach, Charles N. Arge

Previous ambient solar wind (SW) validation studies have reported on discrepancies between modeled and observed SW conditions at L1. They indicated that a major source of discrepancies stems from how we model the solar corona. Thus, enhancing predictive capabilities demands a thorough examination of coronal modeling. The Wang-Sheeley-Arge (WSA) model has been a workhorse model that provides the near-Sun SW conditions. An important component of it is the Potential Field Source Surface (PFSS) model. This study analyzes 15 different Carrington Rotations(CRs), and presents detailed analysis of CR 2052 to identify WSA model settings that lead to successful and erroneous SW predictions at Earth. For the events studied, we show that increasing the models grid resolution improves the open-close boundary identification. This results in better predicting the onset and duration of high-speed streams (HSSs). In addition, we find an optimized source surface height (R$_{ss}$) (lying between 1.8-3.1 R$_{\odot}$) further enhances HSS prediction accuracy for the studied events. A detailed analysis shows that changes in R$_{ss}$, (a) changes the Great Circle Angular Distance (GCAD) maps (at the solar surface) of the associated coronal holes and (b) changes the foot-point locations of the magnetic connectivities to the sub-Earth locations. These factors change the near-Sun SW speed, that eventually leads to uncertainties in speeds near Earth. We also investigate the usefulness of coronal hole observations in constraining R$_{ss}$ and SW solutions at Earth, and highlight their underutilized value in guiding the selection of magnetic maps for improved ambient solar wind modeling at L1.