This is what undertook Véronique Bommier and Guillaume Molodij, of Paris-Meudon Observatory, while observing with an extreme sensitivity the polarization with the Franco-Italian solar telescope THEMIS during its third observing campaign, in august 2000. THEMIS is settled on the site of the Teide volcano, of the "Instituto de Astrofísica de Canarias" (IAC), on the Tenerife Island (Canarias, Spain). It is a telescope devoted to solar observation, and which has been especially designed for polarization measurements: the instrumental polarization has been reduced to a minimum by setting the polarization analyzer before any oblique reflection. Such a telescope is called a "polarization free" telescope.
The radiation polarization is sensitive to the magnetic field: thus, the scientific objective of THEMIS is the magnetic field measurement, by interpretation of the polarization measurements. The circular polarization is rather sensitive to the "longitudinal" field, i.e., the field component along the line-of-sight, whereas the linear polarization is rather sensitive to the "transverse" field, i.e., the field components perpendicular to the line-of-sight. THEMIS measures both polarizations, circular and linear (4 Stokes parameters), aiming to derive maps of the vector magnetic field of the Sun’s surface, with a high spatial resolution.The observations presented here aim to reach the highest possible polarimetric sensitivity with THEMIS: to this purpose, the spatial resolution is degraded by averaging along the spectrograph entrance slit and a large number of images are added in time. Thus has been reached a polarimetric sensitivity of a few 10-5.A high polarimetric sensitivity is required to observe the so-called "second solar spectrum", which is the spectrum of the polarization formed by scattering very close to the solar limb. This polarization is weak, as it can be seen in Figure 2. These observations have been made with the spectrograph entrance slit parallel to the solar limb, 4 arcsec inside the disk, in a quiet solar region: the solar North Pole. When observing near the solar limb, the lines are linearly polarized, due to the scattering of the underlying anisotropic radiation. This anisotropy results from the radiative transfer near the surface of the Sun, where the lines are formed. The observed lines are the Sodium D1 (right hand-side) and D2 (left hand-side ) lines, which are two very deep absorption lines of the intensity spectrum (in purple in the Figur e). Very similar in the intensity spectrum, these two lines are on the contrary very different in the po larization spectrum (in white in the Figure), because the D2 (left hand-side) line is polarizable, w hereas the D1 (right hand-side) line is unpolarizable. The observational results show a polarization spectrum that differs from those predicted by the recent theoretical models, in particular in the D1 component which is known as unpolarizable, whereas polarization peaks are observed. As for the D2 component, the polarization may have been modified by a weak magnetic field, which can depolarize the line (through what is called Hanle effect). But then, the polarization should rotate, i.e. it should change direction, which is not observed. A solution might be that the magnetic field is turbulent, with no precise orientation. But here too, the implications of this hypothesis are not compatible with all polarisation measurements, for all lines, and all atoms (such as strontium or baryum) which are as many constraints on models. Some authors propose that the observed spectrum in the sodium lines is mainly due to the lower levels atomic polarization (with hyperfine structure), coupled to coherent scattering; Others estimate that the lower levels atomic polarization may explain the D1 spectrum, but that the D2 spectrum can be interpreted by coherent scattering only: indeed, some calculations in progress at Paris Observatory (PhD thesis of B. Kerkeni, managed by N. Feautrier and A. Sielpfiedel), show that the collisions of the Sodium atom with the surrounding Hydrogen atoms could destroy the lower levels atomic polarization. In order to progress in the understanding and interpretation of this spectrum and of the magnetic field effects, it would then be necessary to achieve a better modelization of the line polarization formation, by solving the coupled equations of polarized radiative transfer and statistical equilibrium of the multilevel atom, taking also into account coherent scattering, which has never been completely achieved for the moment.
Last update on 21 December 2021