A new method for instrumental profile reconstruction of high resolution spectrographs

TL;DR

This paper introduces new Gaussian Process regression methods for reconstructing the instrumental profile of high-resolution spectrographs using Laser Frequency Comb data, significantly improving wavelength calibration accuracy and reducing distortions.

Contribution

Developed novel Gaussian Process-based methods for detailed IP reconstruction in high-resolution spectrographs with LFC calibration, demonstrating improved wavelength accuracy and reduced distortions.

Findings

Empirical IP models achieve wavelength accuracy better than 10 m/s.

Assuming Gaussian IP shape leads to significant wavelength distortions.

HARPS IP is asymmetric and varies smoothly across the detector.

Abstract

Knowledge of the spectrograph's instrumental profile (IP) provides important information needed for wavelength calibration and for the use in scientific analyses. This work develops new methods for IP reconstruction in high resolution spectrographs equipped with Laser Frequency Comb calibration (LFC) systems and assesses the impact that assumptions on IP shape have on achieving accurate spectroscopic measurements. Astronomical LFCs produce $\approx10000$ bright, unresolved emission lines with known wavelengths, making them excellent probes of the IP. New methods based on Gaussian Process regression were developed to extract detailed information on the IP shape from this data. Applying them to HARPS, an extremely stable spectrograph installed on the ESO 3.6m telescope, we reconstructed its IP at 512 locations of the detector, covering 60% of the total detector area. We found that the HARPS IP is asymmetric and that it varies smoothly across the detector. Empirical IP models provide wavelength accuracy better than 10 ms$^{-1}$ (5 ms$^{-1}$) with 92% (64%) probability. In comparison, reaching the same accuracy has a probability of only 29% (8%) when a Gaussian IP shape is assumed. Furthermore, the Gaussian assumption is associated with intra-order and inter-order distortions in the HARPS wavelength scale as large as 60ms$^{-1}$. The spatial distribution of these distortions suggests they may be related to spectrograph optics and therefore may generally appear in cross-dispersed echelle spectrographs when Gaussian IPs are used. Methods presented here can be applied to other instruments equipped with LFCs, such as ESPRESSO, but also ANDES and G-CLEF in the future. The empirical IPs will be crucial for obtaining objective and unbiased measurements of fundamental constants from high resolution spectra, as well as measurements of the redshift drift, isotopic abundances, and other science cases.