Gaia Data Release 3

Abstract

Context. Diffuse interstellar bands (DIBs) are common interstellar absorption features in spectroscopic observations but their origins remain unclear. DIBs play an important role in the life cycle of the interstellar medium (ISM) and can also be used to trace Galactic structure. Aims. Here, we demonstrate the capacity of the Gaia-Radial Velocity Spectrometer (RVS) in Gaia DR3 to reveal the spatial distribution of the unknown molecular species responsible for the most prominent DIB at 862 nm in the RVS passband, exploring the Galactic ISM within a few kiloparsecs from the Sun. Methods. The DIBs are measured within the GSP-Spec module using a Gaussian profile fit for cool stars and a Gaussian process for hot stars. In addition to the equivalent widths and their uncertainties, Gaia DR3 provides their characteristic central wavelength, width, and quality flags. Results. We present an extensive sample of 476 117 individual DIB measurements obtained in a homogeneous way covering the entire sky. We compare spatial distributions of the DIB carrier with interstellar reddening and find evidence that DIB carriers are present in a local bubble around the Sun which contains nearly no dust. We characterised the DIB equivalent width with a local density of 0.19 ± 0.04 Å/kpc and a scale height of 98.60 +11.10 −8.46 pc. The latter is smaller than the dust scale height, indicating that DIBs are more concentrated towards the Galactic plane. We determine the rest-frame wavelength with unprecedented precision (λ 0 = 8620.86 ± 0.019 Å in air) and reveal a remarkable correspondence between the DIB velocities and the CO gas velocities, suggesting that the 862 nm DIB carrier is related to macro-molecules. Conclusions. We demonstrate the unique capacity of Gaia to trace the spatial structure of the Galactic ISM using the 862 nm DIB.

Measuring the 8621 A Diffuse Interstellar Band in Gaia DR3 RVS Spectra: Obtaining a Clean Catalog by Marginalizing over Stellar Types

“…The primary Gaia analysis of these spectra falls under the General Stellar Parameteriser-spectroscopy (GSP-Spec) module, which is a purely spectroscopic analysis to obtain chemophysical stellar parameters from combined RVS spectra (Recio-Blanco et al 2022). Within GSP-Spec, DIB detections and parameter estimates were made predominantly by dividing each RVS spectrum by the best-matching synthetic spectrum (using the GSP-Spec atmospheric parameters) and performing a least-squares fit of a Gaussian function (Schultheis et al 2023).…”

“…Based on the full 5,591,594 spectra, the Gaia DIB catalog has entries for 476,532 spectra, of which 50,787 have public spectra (Table 1). The Gaia DIB team strongly suggests restricting this catalog to a "highquality" (HQ) sample, which includes 137,536 spectra, of which 9763 are public (Schultheis et al 2023). While the Gaia DIB catalog entries are available for all stars-even those without public spectra-we will restrict our comparisons with the Gaia DIB catalog to stars with public spectra, unless otherwise specified.…”

“…Additional cuts (Schultheis et al 2023) are imposed to remove low-quality detections, often as a result of poor stellar modeling. The GSP-Spec flags 1-13, which indicate the goodness of fit for the stellar modeling, are required to be <2 (reasonably good).…”

“…which they were detected. This has been done in the NIR for the 15273 A ̊DIB in APOGEE (Zasowski et al 2015), which searched over ∼100 k spectra, and in the optical for the 8621 A DIB in Gaia RVS spectra (Schultheis et al 2023), which searched over ∼5.5M spectra. Yet, statistical methods for modeling and detecting DIBs have not kept pace with the increase in data quantity and quality.…”

“…DIB catalogs are still made by obtaining a best-fit stellar spectrum, dividing the observed spectrum by the stellar model, and fitting a Gaussian peak to the residuals centered near the DIB wavelength (Zasowski et al 2015;Schultheis et al 2023). However, stellar line lists and models often have significant residuals relative to observations, which can cause confusion when fitting or finding DIB features in the residuals.…”