Solar Flux Atlases: The new HARPS-N Quiet Sun Benchmark and Ca II H & K Lines Continuum Normalisation

TL;DR

The paper introduces HARPS-N QS as an absolute quiet-Sun reference and compares solar activity across major flux atlases using equivalent widths, radial velocities, and a novel activity metric. It shows that instrumental effects dominate line profiles and intensities, with Ca II H&K normalisation introducing substantial shifts, and demonstrates that MA (measured activity) data have much lower apparent activity than others. The QS atlas provides a stable standard for cross-instrument calibration and solar-state benchmarking, while intra-atlas activity tracking requires instrument-consistent references. The findings underscore the need for detailed observation logs and instrument-matched quiet-Sun atlases to enable high-precision solar-stellar comparisons.

Abstract

Context. Solar flux atlases observe the spatially integrated light from the Sun, treating it as a star. They are fundamental tools for gaining insight into the composition of the Sun and other stars, and are utilized as reference material for a wide range of solar applications such as stellar chemical abundances, atmospheric physics, stellar activity, and radial velocity signals. Aims. We provide a detailed comparison of solar activity in some of the well-known solar atlases against the new High Accuracy Radial velocity Planet Searcher for the Northern hemisphere (HARPS-N) Quiet Sun (QS) and measured activity (MA) atlases that are published, for the first time, in this work. Methods. Ten of the widely used individual spectral lines from each flux atlas were selected to compare solar activity using three methods: 1) Equivalent Widths 2) Activity number, a novel activity measure which we introduce in this work. 3) Bisectors and radial velocity. Results. The significantly smaller activity measured in the MA atlas compared to the other atlases, relative to the QS atlas, underscores the dominance of instrumental effects over solar activity in their impact on spectral lines, which cannot be corrected through simple line convolution to match resolutions of other atlases. Additionally, our investigation unexpectedly revealed a substantial intensity shift in the Ca ii H & K lines of other atlases compared to our HARPS-N atlases, likely caused by assumptions in normalisation techniques used in the early Kitt Peak atlases. Conclusions. With an average spot number of zero, our QS atlas is well suited to serve as an absolute benchmark atlas representative of solar minimum for the visible spectrum that other atlases can be compared against. Our recommendations are 1) publication of a detailed log along with the observations, to include exact dates and indications of solar activity...

Figures (18)

• Figure 1: Solar activity over time as measured by the total filling factor of active regions (see main text). The observation time for each of the atlases compared in this study are shown to illustrate the average level of activity. The atlases are neckel1987spectral, Wallace2011ApJS..195....6W, Molaro2013AA...560A..61M, Reiners2016, Strassmeier2018AA...612A..44S, HARPS-N's MA & QS atlases, and Ellwarth2023, respectively. No information is provided regarding the observing dates of the Meftah2023RemS...15.3560M atlas.
• Figure 2: Overview of atlas activity. Each disc on the top row represents the most active, with the highest spot number, day used for said atlas. The bottom row is the same except that it shows the lowest activity level for any given atlas. Data from the Solar Dynamics Observatory Atmospheric Imaging Assembly Lemen2012SoPh..275...17L 1700 Å filter. Left: HARPS-N quiet Sun atlas, Reiners IAG flux atlas, PEPSI 2018 atlas, and Molaro LFC atlas. We note that the LFC atlas only has one data point, thus, it shows the same data twice.
• Figure 3: Illustrations of the solar disc during the times of the observations for the Wallace flux atlas, observed by the Meudon spectroheliogram Malherbe2019Malherbe23. The images shown are taken in the Ca II K line core, for the times when the atlas observations were taken for each wavelength band, which makes-up the complete Wallace flux atlas. The images contain observational artefacts seen as nearly vertical stripes, most likely caused by clouds.
• Figure 4: Sunspot drawings in white light, made at the Specola Solare Ticineseduring the observations for the Neckel flux atlas. We use the same image to represent the solar disc for several combined wavelength bands that are separated in the atlas. In those cases, the observations were taken within one day of each other and, thus, the disc image is reasonably representative of the solar disc for all those bands.
• Figure 5: Illustration of how the activity number method transforms the intensity values of the spectral lines of other atlases. Top panels: the transformed spectral lines after being convolved and interpolated. The two vertical lines indicate the wavelength region used to calculate the activity number. Bottom panels: the remainder of the ratio differences, as described in Section \ref{['subsec: an']}. For Meftah2023RemS...15.3560M, Reiners2016 and Wallace2011ApJS..195....6W atlases, the wavelengths were converted from vacuum to air using the formula from Birch1994.