日出三号观测中宁静太阳网络内的水平磁场占主导：来自传统反演与基于Transformer的反演的证据
Horizontal Magnetic Fields Dominate the Quiet Sun Internetwork in Sunrise III Observations. Evidence from Traditional and Transformer-Based Inversions
浏览论文内容
中文总结 AI 辅助
该研究针对宁静太阳网络内水平与垂直磁场占比的争议，通过将基于Transformer的SINN反演与传统DeSIRe反演应用于Sunrise III/SCIP观测数据，证实宁静太阳网络内水平磁场显著占优，且基于Transformer的反演可高效迁移至真实观测。
中文摘要 AI 辅助
宁静太阳网络内水平磁场与垂直磁场的相对占比仍存在争议，原因在于弱线性偏振信号会使推断的磁倾角分布对观测质量和反演方法敏感。我们对{\textsc{Sunrise~iii}}观测中的宁静太阳网络磁拓扑展开研究，将基于Transformer的反演技术与传统反演技术独立应用于同一光谱偏振数据集，分析由日出色球层红外光谱偏振仪（SCIP）获取的观测数据。仅在大型辐射磁流体动力学模拟生成的合成斯托克斯轮廓上训练的基于Transformer的反演方法（SINN），与使用\textsc{DeSIRe}代码完成的独立反演进行对比。两种反演方法均得出宁静太阳磁拓扑：网络内的磁场以水平分量为主，而网络外的磁场拓扑则更为均衡。在$\boldsymbol{\rm log}\tau=-1.2$处，采用\textsc{DeSIRe}得到的网络内水平磁场和无符号视线磁场分量的中位数分别为64G和7G，采用SINN得到的对应值分别为71G和8G，两种方法对应的水平-视线分量比值均约为9。两种根本不同的反演方法的一致性表明，宁静太阳网络的特征是水平磁场分量显著强于垂直磁场分量。这些结果证明，基于Transformer的反演可成功从合成训练数据迁移至真实光谱偏振观测，且其推理速度比传统反演技术快数个数量级。
英文摘要
The relative prevalence of horizontal and vertical magnetic fields in the quiet-Sun internetwork remains debated, owing to the weak linear polarisation signals that make the inferred magnetic inclination distributions sensitive to observational quality and inversion methodology. We investigate the magnetic topology of the quiet-Sun internetwork in {\sc Sunrise~iii} observations by independently applying transformer-based and traditional inversion techniques to the same spectropolarimetric dataset. We analyse observations obtained with the Sunrise Chromospheric Infrared spectroPolarimeter (SCIP). Transformer-based inversions (SINN), trained exclusively on synthetic Stokes profiles generated from a large radiative magnetohydrodynamic simulation, are compared with independent inversions performed using the \textsc{DeSIRe} code. Both inversion methods recover a quiet-Sun magnetic topology in which internetwork fields are strongly dominated by the horizontal component, whereas network fields exhibit a substantially more balanced topology. At $\logτ=-1.2$, the median horizontal and unsigned line-of-sight field components in the internetwork are 64 and 7G with \textsc{DeSIRe}, and 71 and 8G with SINN, respectively, corresponding to a horizontal-to-line-of-sight ratio of approximately nine for both methods. The agreement between the two fundamentally different inversion methodologies indicates that the quiet-Sun internetwork is characterised by a substantially stronger horizontal than vertical magnetic field component. These results demonstrate that transformer-based inversions can be successfully transferred from synthetic training data to real spectropolarimetric observations, while providing inference orders of magnitude faster than traditional inversion techniques.
