Image of the Week

How Gaia unveils what stars are made of

Several Gaia RVS Spectra, presenting the stellar flux (black dots) in the wavelength interval between 846 and 870 nm (in vacuum). The coloured vertical stripes indicate the different atoms and molecules present in the stellar atmosphere. A diffuse interstellar absorption band (DIB) is highlighted by a green square. Credit: ESA/Gaia/DPAC, CC BY-SA 3.0 IGO. Acknowledgement: Gaia/DPAC/CU8-CU6, Recio-Blanco and the GSP-Spec team.

Gaia RVS stellar spectra, at a resolution of ~11500, contain a wealth of information about Milky Way stars. This animation shows several Gaia RVS stellar spectra, presenting the stellar flux (black dots) in the wavelength interval between 846 and 870 nm (in vacuum). The flux is normalized to the continuum, showing variations (spectral lines) due to the light absorption from atoms and molecules present in the stellar atmosphere, that are identified by coloured vertical stripes in the figure.

These absorption features, and therefore the spectral line profiles, depend on the physical properties of the star. In particular, this dependence carries information on the effective temperature (T_eff, in K), the surface gravitational acceleration (usually expressed through its logarithm, log g, with g in cm/s²), the mean abundance of elements heavier than helium ([M/H] in dex) and the mean abundance of α-elements (oxygen, magnesium, silicon, sulphur, calcium, titanium) with respect to iron ([α/Fe]), expressed in dex, a scientific notation converting the number preceding it into its 10 based antilogarithm. In addition, the absorption increases with each element’s or molecule’s abundance, allowing to estimate the chemical composition of the stellar atmosphere.

The Gaia Generalized Stellar Parameterizer-Spectroscopy (GSP-Spec) algorithm of the Apsis chain of Coordination Unit 8 in the Gaia Data Processing and Analysis Consortium is able to estimate up to 18 different physical parameters and chemical abundances for millions of RVS spectra. Processing of the Gaia data to produce these astrophysical parameters has been recently finalized and validation of the data is currently ongoing, with the expected release of the results with Gaia Data Release 3 in the first half of 2022.

First of all, GSP-Spec provides the four atmospheric parameters (T_eff, log g, [M/H], [α/Fe]) that are reported on top of the image. In addition, around 60 different spectral lines are automatically analysed and then combined to derive the abundance of up to 12 chemical species, as shown on the left side of the image. For each element, the absorption lines detection, conditioned to their depth, is highly dependent on the atmospheric parameters and the spectra quality (such as the signal-to-noise ratio, chosen to be very high for the spectra illustrated in the animation). This detection determines the possibility of estimating the corresponding abundance in the stellar atmosphere. As a consequence, not all the elements are estimated for all the spectra, with important variations from one element to another. For each element, if at least one line is detected, an estimation of the corresponding abundance in the stellar atmosphere becomes possible. If several lines of the same element are detected, their individual results are combined, taking into account their quality (uncertainties).

All chemical abundances, including global metallicity and [α/Fe], are expressed with respect to the Solar value, in a logarithmic scale (e.g. [Fe/H]=-0.30 dex indicates that the stellar iron abundance is half of the Solar one). Moreover, all the displayed chemical abundances are expressed with respect to the iron one, with the exception of iron itself, whose abundance is derived with respect to hydrogen.

Finally, the intensity (equivalent width, EW) and central position (λ) of a diffuse interstellar absorption band (DIB) highlighted by a green square in the animation above, are also provided by the GSP-Spec module. DIBs are interstellar absorption features originating from the interstellar medium, quasi-consensually attributed  to large organic molecules (see also: Diffuse Interstellar Bands in three hot stars as seen by Gaia's Radial Velocity Spectrometer)

The chemical composition of the stellar atmosphere informs us on the interstellar medium conditions at the time of the star's formation, a crucial fossil signature to reconstruct the history of our Galaxy. Gaia chemical diagnostics include all the main nucleosynthetic channels enriching the interstellar medium, from exploding massive stars, Type Ia supernovae, neutron-star mergers and dying low mass stars.

The RVS spectra, calibrated by the CU6 pipeline and analysed by the GSP-Spec module of the CU8 Apsis pipeline, will be published in 2022 as part of the Gaia Data Release 3.

Abbreviations:

• DPAC = Data Processing and Analysis Consortium
• CU6 = Coordination Unit 6
• CU8 = Coordination Unit 8
• RVS = Radial Velocity Spectrometer
• GSP-Spec = Gaia Generalized Stellar Parameterizer-Spectroscopy
• Apsis = Gaia Astrophysical Parameters Inference System

Credits: ESA/Gaia/DPAC/CU8-CU6, Alejandra Recio-Blanco (OCA), Patrick de Laverny (OCA), Georges Kordopatis (OCA), Christophe Ordenovic (OCA), Pedro Alonso-Palicio (OCA), Mathias Schultheis (OCA), Marco Antonio Alvarez-González (GGG), He Zhao (OCA), Gabriele Contursi (OCA), Minia Manteiga (GGG), Carlos Dafonte (GGG), Inna Oreshina-Slezak (OCA). We wish to thank the Gaia Data Processing Centre at the Centre National d'Etudes Spatiales (CNES; DPCC) for producing the high-quality spectroscopy and data analysis upon which this work rests.

[Published: 09/07/2021]