|dc.description.abstracteng||In this thesis we investigate small-scale scattering polarization signals in quiet, internetwork (IN) regions of the lower solar atmosphere.
Small-scale polarization signals are inherently hard to observe due to their small spatial scale structuring with low amplitude and probably fast changing nature. We aim to measure and understand the horizontal fluctuations of these polarization signals, as this is the first step towards spatially resolved Hanle effect observations. Hanle effect observations are a promising complementary diagnostic tool for highly tangled and dynamic magnetic fields, which are invisible to standard Zeeman effect based observations. Therefore, Hanle effect observations have the potential to shed light on the small-scale dynamo action on the Sun's surface. But even if the scattering signals are not explicitly interpreted in terms of the Hanle effect, physical insights can still be gained into the formation and modification of the scattering polarization.
We use high spatial and temporal resolution observations, obtained by two ground-based polarimeters: the Fast Solar Polarimeter prototype (FSP) and an updated and modified version, FSP 2, to address the challenges of high polarimetric sensitivity in combination with high spatio-temporal resolution. We concentrate on the statistical analysis of the scattering polarization of the prominent Sr I spectral line in magnetically quiet solar regions. Some part of the analysis is performed using a novel method we developed to classify image pixels and average them according to the inhomogeneity of the solar granulation. This inhomogeneity breaks the axial symmetry of the radiation field, which was predicted to cause scattering polarization at solar disk center. Within this context, we have made an original contribution to knowledge in three areas: We produced the first published filtergraph observation of scattering polarization in Sr I, where we found an anti-correlation between linear polarization and continuum intensity; we found the first significant evidence of spatially structured scattering polarization in Sr I at disk center; and we developed a new tool which opens possibilities to analyze observations of scattering polarization, even in low signal-to-noise regimes and compare them to numerical models.
In the first part of the thesis, we give an overview of the small-scale turbulent magnetic field on the Sun and polarizing mechanisms in the solar photosphere. Within the general theoretical framework of polarized spectral lines provided by Landi Degl'Innocenti (2004) we focus on how scattering polarization emerges from anisotropic radiation and how this can be formally described. In doing so, we will not only cover the well known case for scattering polarization of spectral lines at the solar limb, but also at solar disk center.
To motivate how scattering polarization can be applied in astrophysical diagnostics, we briefly elaborate on the difference between the two main magnetic field diagnostic techniques, the Hanle and the Zeeman effect. We discuss why the Hanle effect, which modifies scattering polarization, is more suitable for diagnosing small-scale turbulent magnetic fields.
In the second part of the thesis, we discuss a filtergraph observation at the German Vacuum Tower Telescope on Tenerife carried out with the FSP.
We present the observational evidence of spatially structured scattering polarization. The scattering polarization is structured with respect to intergranules and granules. Our statistical analysis reveals that the linear polarization component parallel to the north solar limb in the Sr I line core anti-correlates with the continuum intensity. Furthermore, we found that the spatial dimension of these structures are on the order of 0.5"-1". We show that the polarization signals are consistent with numerical models. These models suggest that the polarization predominantly emerges from intergranules, i.e. more precisely, from the interface between intergranules and granules, due to local radiation anisotropy. This finding, however, contradicts the result of two studies that were conducted earlier (Malherbe et al. 2007, Bianda et al. 2018). Especially these studies were carried out with spectrographs and at a solar limb distance of µ=0.3, while our filtergraph observation was done at µ=0.6.
In the third part of the thesis, we study the linear polarization signals in the Sr I line at solar disk center observed with FSP 2 at the Dunn Solar Telescope, New Mexico. Due to geometric considerations, the mean scattering polarization should be zero. However, we found polarization signals which correlate with the local axial symmetry break of the radiation field. As a zero reference, we observed the continuum and a neighboring Fe I spectral line position. The latter spectral line serves as a zero reference, as it is insensitive to scattering. We introduce a statistical method that allows for the first time to show horizontally fluctuating scattering polarization in both linear polarization states. We are even able to convert the statistical values back to images and roughly reconstruct the linear polarization maps. As we used a medium sized solar telescope, this opens the unexpected opportunity to study these signals already in currently available solar observatories, where signal-to-noise ratios are not sufficient to detect such small-scale scattering signals directly.
These observational results support a range of theoretical predictions (del Pino Aleman et al. 2018). Our findings support the picture of an isotropic magnetic field in the quiet Sun, with magnetic field amplitudes small enough to safely stay away from the Hanle saturation regime (~100 G for Sr I). Our analysis furthermore supports and further restricts the findings from the second part of the thesis. We find that the mean spatial size for scattering polarization in this observation is about 0.75", but larger structures (~2") are visible, too.
Finally, we compare the results found in the third part with a published state-of-the-art 3D-MHD simulation snapshot, where the Sr I line has been synthesized using the 3D radiative transfer code PORTA (del Pino Aleman at al. 2018). The magnetic field and collisional rates have been chosen to reproduce center-to-limb observations of Sr I. We degrade the synthesized data to the FSP 2 observations. Our analysis of the comparison between the simulation and the observation reveals very interesting findings on two levels. Firstly, we are able to test the statistical method introduced in the second part by applying it to noisy and noise-free simulation data and find that it is more reliable for structures larger than 1". This preliminary result implies that the size estimation in the third part of this thesis may be biased towards larger structures. Secondly, the spatial distribution of the scattering polarization in the reconstructed observation is comparable to the simulation. However, the polarization amplitude in the observation is reduced by a factor of two compared to the simulation. We find indications that temporal evolution of the Sun is the main cause of this reduction of the observed scattering polarization amplitude, not additional magnetic fields. This may indicate that the simulated photosphere is magnetized in the right amount to mimic the "true" photosphere, which means that the solar lower atmosphere may be magnetized close to equipartition. Two other (but less probable) possibilities are that either the depolarizing collision rates are much higher than expected by center-to-limb observations of Sr I or/and that the small-scale, yet undetected magnetic field is even stronger and/or more dynamic in the Sun than is suggested by center-to-limb observations of Sr I. The latter case, if it proves to be true, implies that there is more magnetic energy hidden in the Sun than was previously expected.||de