On the role of source surface height and magnetograms in solar wind forecast accuracy

Many operational space weather forecasting frameworks are based on the Potential Field Source Surface (PFSS) model of the magnetic field. The output of PFSS serves as input in many heliospheric models that provide solar wind velocity predictions at L1. Previous studies in the context of prediction of open magnetic flux observed at L1 have suggested different source surface heights (Rss) for the PFSS model at different phases of the solar cycle (SC). We investigate the effects and necessity of optimizing the Rss in the PFSS model in the context of its use in the popular Wang-Sheeley-Arge (WSA) model for solar wind velocity prediction. We used Heliospheric Upwind Extrapolation (HUX) to extrapolate solar wind velocity in the heliosphere. We performed a study of 16 Carrington Rotations (CR) at different phases of the SC24 and SC25, using different types of magnetograms and WSA model parameters. We combine the coronal models (PFSS+WSA) with the heliospheric model (HUX) to predict solar wind velocity at L1 in our framework, i.e., PFSS+WSA+HUX. Our study suggests that using a higher Rss (3.0 R⊙) compared to the conventional Rss (2.5 R⊙) near the solar minimum, results in an improvement in the average Pearson’s correlation coefficient (cc) from 0.61 to 0.75 between the observed and modeled values of solar wind velocity profile at L1. We found that the performance of the framework improved by using zero-point corrected (ZPC) maps in comparison to the standard (STD) Carrington maps from GONG, as demonstrated by an increase in the correlation coefficient from 0.31 to 0.51. We also found that the improved performance of the framework for ZPC maps as compared to the STD full Carrington maps, can be attributed to its capability to capture the global magnetic field. This was further confirmed by comparing the extrapolated global magnetic field structures with the large-scale corona observed in the extended field of view of the PROBA2/SWAP images. Our work is a first step in the direction of improving the WSA model and points out the potential ways to enhance the PFSS+WSA framework of solar wind forecasting at L1.