Magnetic connectivity from the Sun to the Earth with MHD models II. Impact of input magnetic maps

Abstract

Connectivity estimation needs to be fast and thus usually relies on semi-empirical models. We previously established that optimised magnetohydrodynamics (MHD) models can also be used to derive magnetic connectivity, yielding good results, especially for complex configurations. However, the modelling is not the only factor that can impact the connectivity estimation. Models' results, both semi-empirical and MHD, depend heavily on the input data used to drive the simulation. In order to quantify this impact on our connectivity estimation, we used 20 different maps by four providers. We selected the period of July 2019, which corresponds to Carrington rotation (CR) number 2219, because it had already been studied in a previous paper. It corresponds to a period of minimum activity, meaning minimal interference from transient structures, but with a high-speed stream and several polarity switches. We have compared our results with in situ data from ACE and developed several visualisation techniques to validate connectivity not only for a specific date but also for a full CR. The choice of the input magnetic map drastically affects the connectivity estimates, with global scores going from 28% global agreement to 63% in the most extreme case. The HMI and GONG maps consistently gave the best connectivity estimations, with more than 50% correct results of the full CR. For this specific period, the GONG-ADAPT maps yielded the least reliable results, with a score below 35%. This is not surprising, as we have shown before that COCONUT required parameter adjustments for these maps. However, WSO maps which showed poor coronal agreement actually provided good connectivity estimates, showing that the Heliospheric Current Sheet (HCS) location is in the end the most crucial parameter for connectivity. Consistent data over a full Carrington rotation are to be favored compared to daily.