Second Solar Spectrum

• Gandorfer, A. 2000, The Second Solar Spectrum: A high spectral resolution polarimetric survey of scattering polarization at the solar limb in graphical representation. Volume I: 4625 Å to 6995 Å. Zurich: VdF
• Gandorfer, A. 2002, The Second Solar Spectrum: A high spectral resolution polarimetric survey of scattering polarization at the solar limb in graphical representation. Volume II: 3910 Å to 4630 Å. Zurich: VdF
• Gandorfer, A. 2005, The Second Solar Spectrum: A high spectral resolution polarimetric survey of scattering polarization at the solar limb in graphical representation. Volume III: 3160 Å to 3915 Å. Zurich: VdF
• Stenflo, J.O. 2014, in R. Ramelli (ed.), http://www.irsol.usi.ch/data/data_archive/#ss2

The full atlas (3161.2 – 6987.0 Å) has been assembled (by J.O. Stenflo) from a number of digital sections provided by Achim Gandorfer, which he used for the three volumes of his atlas. In the present reassembly of the atlas (into a single atlas file) the different sections with arbitrary zero points of the polarization scale have been shifted to be consistent with the empirical determination of the continuum polarization in the Stenflo (A&A 429, 713-730, 2005) paper. The level of this continuum polarization is overplotted as a dashed line.

Above 4000 Å it is possible to obtain a well defined continuum polarization level and zero point of the scale this way. Below 4000 Å it is much more problematic because the spectrum is so crowded with no good continuum windows. Therefore no great effort has been invested in trying to define the zero point of the polarization scale in a precise way. The choices made in the atlas should only be seen as crude indications, with room for considerable refinement if one wants to study a particular section in more detail. For the range (over more than 100 Å) covered by the Ca II H and K lines the zero points have been chosen to make the atlas consistent with the recordings published in Stenflo (A&A 84, 68-74, 1980) (which covered a large spectral range in one context, in contrast to the ZIMPOL recordings, which represent a set of separate small (5-6 Å) spectral windows with individual and arbitrary zero points, which need to be pieced together to achieve large spectral coverage).

There are two versions of the full atlas, represented by two pdf files:

• SSSatlas0.pdf: unsmoothed, as in the original version of Gandorfer.
• SSSatlas.pdf: wavelet-smoothed version.

A very conservative, mild form of wavelet smoothing has been used, to avoid the risk of generating spurious features or significantly affect the polarization amplitudes. The smoothed version is cleaner and better and the one recommended for use, but the unsmoothed version is provided to allow comparison and verification that there is full consistency between the two versions. No inconsistency between the two versions has yet been identified. If someone detects an inconsistency, please report it (preferably to Stenflo, who is responsible for the wavelet smoothing).

The pdf atlas is presented in the form of 10 Å sections. To optimize the visibility of the structures, the Q/I scale has been maximized for each 10 Å section, between the maximum value and the smaller of zero and the minimum value. This is in contrast to the three printed volumes of the original Gandorfer atlas, which used fixed Q/I scales. A particular consequence of this is that in the long-wavelength portion of the atlas, where the polarization signals are very small, the signals including the noise get greatly magnified. This brings out tiny but still significant polarization features which are hardly noticeable in the printed atlas version, while also magnifying spurious features, like polarized fringes and noise, which occur at levels of a few times 10^-5. In general features with Q/I < 3 · 10^-5 should not be trusted, but larger amplitudes are most likely significant.

The UV portion 3165.0 – 4231.0 Å is also covered by the old Kitt Peak survey of the Second Solar Spectrum made on October 6-8, 1978, with the vertical grating spectrometer at Kitt Peak/McMath and published in Stenflo et al. (A&A Suppl. Ser., 52, 161-180, 1983). The spectral resolution was low (in contrast to the fully resolved FTS spectra), about 0.1 Å, determined by the use of relatively wide entrance and exit slits to optimize polarimetric sensitivity. Nevertheless this low-resolution atlas is still of interest since it has a better definition of the continuum level and less fluctuations in the zero polarization level, since the recordings were made through scanning with the grating over 185 Å wide spectral sections, while the ZIMPOL atlas of Gandorfer was pieced together from spectral windows of 5-6 Å width each. The slit position was at 10 arcsec inside the limb, corresponding to µ = 0.144 (in contrast to (µ = 0.10 for the Gandorfer ZIMPOL atlas).