Inferring Maps of the Sun’s Far-side Unsigned Magnetic Flux from Far-side Helioseismic Images Using Machine Learning Techniques

Abstract

Accurate modeling of the Sun’s coronal magnetic field and solar wind structures requires inputs of the solar global magnetic field, including both the near and far sides, but the Sun’s far-side magnetic field cannot be directly observed. However, the Sun’s far-side active regions are routinely monitored by helioseismic imaging methods, which only require continuous near-side observations. It is therefore both feasible and useful to estimate the far-side magnetic-flux maps using the far-side helioseismic images despite their relatively low spatial resolution and large uncertainties. In this work, we train two machine-learning models to achieve this goal. The first machine-learning training pairs simultaneous Solar Dynamics Observatory (SDO)/Helioseismic and Magnetic Imager-observed magnetic-flux maps and SDO/Atmospheric Imaging Assembly-observed extreme ultraviolet (EUV) 304 Å images, and the resulting model can convert 304 Å images into magnetic-flux maps. This model is then applied to the STEREO-observed far-side 304 Å images, available for about 4.3 yr, for the far-side magnetic-flux maps. These EUV-converted magnetic-flux maps are then paired with simultaneous far-side helioseismic images for a second machine-learning training, and the resulting model can convert far-side helioseismic images into magnetic-flux maps. These helioseismically derived far-side magnetic-flux maps, despite their limitations in spatial resolution and accuracy, can be routinely available on a daily basis, providing useful magnetic information on the Sun’s far side using only the near-side observations.

The Advective Flux Transport Model: Improving the Far Side with Active Regions Observed by STEREO 304 Å

“…The advent of ML/AI is providing new avenues for advancing helioseismic techniques and improving the accuracy. Both the HMI and GONG helioseismology teams are currently exploring ML/AI methodologies to provide improved data products, including helioseismic inferred farside magnetic maps (Chen et al 2022;Creelman et al 2024).…”

Synoptic maps from two viewpoints

“…Later, the Air Force Data Assimilative Photospheric flux Transport (ADAPT; Arge et al 2010) expanded on the work of Harvey & Worden and their evolving synoptic maps, introducing the Los Alamos National Laboratory data assimilation framework in order to account for data and model uncertainties and thus produce improved synoptic maps. Designed to work with all common solar data, including magnetic approximations from helioseismic far-side data (Arge et al 2013;Lindsey & Braun 2000;Chen et al 2022), ADAPT models globally instantaneous synchronic maps with consistent polar fields, which serve as input for solar wind or coronal modelling.…”