# 20.2.1 astrophysical_parameters

This is the main table containing the 1D astrophysical parameters produced by the Apsis processing chain developed in Gaia DPAC CU8 (see Chapter 11). Additional parameters can be found in table astrophysical_parameters_supp.

Columns description:

solution_id : Solution Identifier (long)

All Gaia data processed by the Data Processing and Analysis Consortium comes tagged with a solution identifier. This is a numeric field attached to each table row that can be used to unequivocally identify the version of all the subsystems that were used in the generation of the data as well as the input data used. It is mainly for internal DPAC use but is included in the published data releases to enable end users to examine the provenance of processed data products. To decode a given solution ID visit https://gaia.esac.esa.int/decoder/solnDecoder.jsp

source_id : Source Identifier (long)

A unique single numerical identifier of the source obtained from gaia_source (for a detailed description see gaia_source.source_id).

classprob_dsc_combmod_quasar : Probability from DSC-Combmod of being a quasar (data used: BP/RP spectrum, photometry, astrometry) (float)

Probability that the object is of the named class. This is the overall probability for this class, computed by combining the class probabilities from DSC-Specmod (which classifies objects using BP/RP spectra) and DSC-Allosmod (which classifies objects using several astrometric and photometric features). It is important to realise that the DSC classes are defined by the training data used, and that this may produce a narrower definition of the class than may be expected given the class name. This is a posterior probability that includes the global class prior, given in the documentation.

classprob_dsc_combmod_galaxy : Probability from DSC-Combmod of being a galaxy (data used: BP/RP spectrum, photometry, astrometry) (float)

Probability that the object is of the named class. This is the overall probability for this class, computed by combining the class probabilities from DSC-Specmod (which classifies objects using BP/RP spectra) and DSC-Allosmod (which classifies objects using several astrometric and photometric features). It is important to realise that the DSC classes are defined by the training data used, and that this may produce a narrower definition of the class than may be expected given the class name. This is a posterior probability that includes the global class prior, given in the documentation.

classprob_dsc_combmod_star : Probability from DSC-Combmod of being a single star (but not a white dwarf) (data used: BP/RP spectrum, photometry, astrometry) (float)

Probability that the object is of the named class. This is the overall probability for this class, computed by combining the class probabilities from DSC-Specmod (which classifies objects using BP/RP spectra) and DSC-Allosmod (which classifies objects using several astrometric and photometric features). It is important to realise that the DSC classes are defined by the training data used, and that this may produce a narrower definition of the class than may be expected given the class name. This is a posterior probability that includes the global class prior, given in the documentation.

classprob_dsc_combmod_whitedwarf : Probability from DSC-Combmod of being a white dwarf (data used: BP/RP spectrum, photometry, astrometry) (float)

Probability that the object is of the named class. This is the overall probability for this class, computed by combining the class probabilities from DSC-Specmod (which classifies objects using BP/RP spectra) and DSC-Allosmod (which classifies objects using several astrometric and photometric features). It is important to realise that the DSC classes are defined by the training data used, and that this may produce a narrower definition of the class than may be expected given the class name. This is a posterior probability that includes the global class prior, given in the documentation.

classprob_dsc_combmod_binarystar : Probability from DSC-Combmod of being a binary star (data used: BP/RP spectrum, photometry, astrometry) (float)

Probability that the object is of the named class. This is the overall probability for this class, computed by combining the class probabilities from DSC-Specmod (which classifies objects using BP/RP spectra) and DSC-Allosmod (which classifies objects using several astrometric and photometric features). It is important to realise that the DSC classes are defined by the training data used, and that this may produce a narrower definition of the class than may be expected given the class name. This is a posterior probability that includes the global class prior, given in the documentation.

classprob_dsc_specmod_quasar : Probability from DSC-Specmod of being a quasar (data used: BP/RP spectrum) (float)

Probability that the object is of the named class. This is the probability from a classifier that uses the BP/RP spectrum (module DSC-Specmod). It is important to realise that the DSC classes are defined by the training data used, and that this may produce a narrower definition of the class than may be expected given the class name. This is a posterior probability that includes the global class prior, given in the documentation.

classprob_dsc_specmod_galaxy : Probability from DSC-Specmod of being a galaxy (data used: BP/RP spectrum) (float)

Probability that the object is of the named class. This is the probability from a classifier that uses the BP/RP spectrum (module DSC-Specmod). It is important to realise that the DSC classes are defined by the training data used, and that this may produce a narrower definition of the class than may be expected given the class name. This is a posterior probability that includes the global class prior, given in the documentation.

classprob_dsc_specmod_star : Probability from DSC-Specmod of being a single star (but not a white dwarf) (data used: BP/RP spectrum) (float)

Probability that the object is of the named class. This is the probability from a classifier that uses the BP/RP spectrum (module DSC-Specmod). It is important to realise that the DSC classes are defined by the training data used, and that this may produce a narrower definition of the class than may be expected given the class name. This is a posterior probability that includes the global class prior, given in the documentation.

classprob_dsc_specmod_whitedwarf : Probability from DSC-Specmod of being a white dwarf (data used: BP/RP spectrum) (float)

Probability that the object is of the named class. This is the probability from a classifier that uses the BP/RP spectrum (module DSC-Specmod). It is important to realise that the DSC classes are defined by the training data used, and that this may produce a narrower definition of the class than may be expected given the class name. This is a posterior probability that includes the global class prior, given in the documentation.

classprob_dsc_specmod_binarystar : Probability from DSC-Specmod of being a binary star (data used: BP/RP spectrum) (float)

Probability that the object is of the named class. This is the probability from a classifier that uses the BP/RP spectrum (module DSC-Specmod). It is important to realise that the DSC classes are defined by the training data used, and that this may produce a narrower definition of the class than may be expected given the class name. This is a posterior probability that includes the global class prior, given in the documentation.

classprob_dsc_allosmod_quasar : Probability from DSC-Allosmod of being a quasar (data used: photometry, astrometry) (float)

Probability that the object is of the named class. This is the probability from a classifier that uses various astrometric and photometric features (module DSC-Allosmod). It is important to realise that the DSC classes are defined by the training data used, and that this may produce a narrower definition of the class than may be expected given the class name. This is a posterior probability that includes the global class prior, given in the documentation.

classprob_dsc_allosmod_galaxy : Probability from DSC-Allosmod of being a galaxy (data used: photometry, astrometry) (float)

Probability that the object is of the named class. This is the probability from a classifier that uses various astrometric and photometric features (module DSC-Allosmod). It is important to realise that the DSC classes are defined by the training data used, and that this may produce a narrower definition of the class than may be expected given the class name. This is a posterior probability that includes the global class prior, given in the documentation.

classprob_dsc_allosmod_star : Probability from DSC-Allosmod of being a star (data used: photometry, astrometry) (float)

Probability that the object is of the named class. This is the probability from a classifier that uses various astrometric and photometric features (module DSC-Allosmod). It is important to realise that the DSC classes are defined by the training data used, and that this may produce a narrower definition of the class than may be expected given the class name. This is a posterior probability that includes the global class prior, given in the documentation.

teff_gspphot : Effective temperature from GSP-Phot Aeneas best library using BP/RP spectra (float, Temperature[K])

Effective temperature (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax (see Section 11.3.3 of the online documentation). This is the median of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value.

teff_gspphot_lower : Lower confidence level (16%) of effective temperature from GSP-Phot Aeneas best library using BP/RP spectra (float, Temperature[K])

Lower confidence level (16%) of effective temperature (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 16th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval.

teff_gspphot_upper : Upper confidence level (84%) of effective temperature from GSP-Phot Aeneas best library using BP/RP spectra (float, Temperature[K])

Upper confidence level (84%) of effective temperature (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 84th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval.

logg_gspphot : Surface gravity from GSP-Phot Aeneas best library using BP/RP spectra (float, GravitySurface[log cgs])

Surface gravity (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax (see Section 11.3.3 of the online documentation). This is the median of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value.

logg_gspphot_lower : Lower confidence level (16%) of surface gravity from GSP-Phot Aeneas best library using BP/RP spectra (float, GravitySurface[log cgs])

Lower confidence level (16%) of surface gravity (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 16th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval.

logg_gspphot_upper : Upper confidence level (84%) of surface gravity from GSP-Phot Aeneas best library using BP/RP spectra (float, GravitySurface[log cgs])

Upper confidence level (84%) of surface gravity (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 84th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval.

mh_gspphot : Iron abundance from GSP-Phot Aeneas best library using BP/RP spectra (float, Abundances[dex])

Decimal logarithm of the ratio of the number abundance of iron to the number abundance of hydrogen relative to the same ratio of solar abundances inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax, assuming source is a single star (see Section 11.3.3 of the online documentation). This is the median of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value.

mh_gspphot_lower : Lower confidence level (16%) of iron abundance from GSP-Phot Aeneas best library using BP/RP spectra (float, Abundances[dex])

Decimal logarithm of the ratio of the number abundance of iron to the number abundance of hydrogen relative to the same ratio of solar abundances inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax, assuming source is a single star (see Section 11.3.3 of the online documentation). This is the 16th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval.

mh_gspphot_upper : Upper confidence level (84%) of iron abundance from GSP-Phot Aeneas best library using BP/RP spectra (float, Abundances[dex])

Decimal logarithm of the ratio of the number abundance of iron to the number abundance of hydrogen relative to the same ratio of solar abundances inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax, assuming source is a single star (see Section 11.3.3 of the online documentation). This is the 84th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval.

distance_gspphot : Distance from GSP-Phot Aeneas best library using BP/RP spectra (float, Length & Distance[pc])

Distance (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax(see Section 11.3.3 of the online documentation). This is the median of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. NB: The actual fit parameter is $\log_{10}d$ and a prior is imposed to ensure a value between [0,5], thus the minimum possible distance is 1 pc and the maximum is 100 kpc.

distance_gspphot_lower : Lower confidence level (16%) of distance from GSP-Phot Aeneas best library using BP/RP spectra (float, Length & Distance[pc])

Lower confidence level (16%) of distance (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 16th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval. NB: The actual fit parameter is $\log_{10}d$ and a prior is imposed to ensure a value between [0,5], thus the minimum possible distance is 1 pc and the maximum is 100 kpc.

distance_gspphot_upper : Upper confidence level (84%) of distance from GSP-Phot Aeneas best library using BP/RP spectra (float, Length & Distance[pc])

Upper confidence level (84%) of distance (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 84th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval. NB: The actual fit parameter is $\log_{10}d$ and a prior is imposed to ensure a value between [0,5], thus the minimum possible distance is 1 pc and the maximum is 100 kpc.

azero_gspphot : Monochromatic extinction $A_{0}$ at 541.4 nm from GSP-Phot Aeneas best library using BP/RP spectra (float, Magnitude[mag])

Monochromatic extinction $A_{0}$ at 541.4 nm (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the median of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. NB: This is the extinction parameter in the adopted Fitzpatrick extinction law (Fitzpatrick 1999, see Section 11.2.3 of the online documentation).

azero_gspphot_lower : Lower confidence level (16%) of monochromatic extinction $A_{0}$ at 541.4 nm from GSP-Phot Aeneas best library using BP/RP spectra (float, Magnitude[mag])

Lower confidence level (16%) of monochromatic extinction $A_{0}$ at 541.4 nm (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 16th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval. NB: This is the extinction parameter in the adopted Fitzpatrick extinction law (Fitzpatrick 1999, see Section 11.2.3 of the online documentation).

azero_gspphot_upper : Upper confidence level (84%) of monochromatic extinction $A_{0}$ at 541.4 nm from GSP-Phot Aeneas best library using BP/RP spectra (float, Magnitude[mag])

Upper confidence level (84%) of monochromatic extinction $A_{0}$ at 541.4 nm (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 84th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval. NB: This is the extinction parameter in the adopted Fitzpatrick extinction law (Fitzpatrick 1999, see Section 11.2.3 of the online documentation).

ag_gspphot : Extinction in G band from GSP-Phot Aeneas best library using BP/RP spectra (float, Magnitude[mag])

Broadband extinction in G band (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the median of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value.

ag_gspphot_lower : Lower confidence level (16%) of extinction in G band from GSP-Phot Aeneas best library using BP/RP spectra (float, Magnitude[mag])

Lower confidence level (16%) of broadband extinction in G band (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 16th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval.

ag_gspphot_upper : Upper confidence level (84%) of extinction in G band from GSP-Phot Aeneas best library using BP/RP spectra (float, Magnitude[mag])

Upper confidence level (84%) of broadband extinction in G band (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 84th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval.

abp_gspphot : Extinction in $G_{\rm BP}$ band from GSP-Phot Aeneas best library using BP/RP spectra (float, Magnitude[mag])

Broadband extinction in $G_{\rm BP}$ band (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the median of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value.

abp_gspphot_lower : Lower confidence level (16%) of extinction in $G_{\rm BP}$ band from GSP-Phot Aeneas best library using BP/RP spectra (float, Magnitude[mag])

Lower confidence level (16%) of broadband extinction in $G_{\rm BP}$ band (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 16th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval.

abp_gspphot_upper : Upper confidence level (84%) of extinction in $G_{\rm BP}$ band from GSP-Phot Aeneas best library using BP/RP spectra (float, Magnitude[mag])

Upper confidence level (84%) of broadband extinction in $G_{\rm BP}$ band (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 84th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval.

arp_gspphot : Extinction in $G_{\rm RP}$ band from GSP-Phot Aeneas best library using BP/RP spectra (float, Magnitude[mag])

Broadband extinction in $G_{\rm RP}$ band (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the median of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value.

arp_gspphot_lower : Lower confidence level (16%) of extinction in $G_{\rm RP}$ band from GSP-Phot Aeneas best library using BP/RP spectra (float, Magnitude[mag])

Lower confidence level (16%) of broadband extinction in $G_{\rm RP}$ band (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 16th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval.

arp_gspphot_upper : Upper confidence level (84%) of extinction in $G_{\rm RP}$ band from GSP-Phot Aeneas best library using BP/RP spectra (float, Magnitude[mag])

Upper confidence level (84%) of broadband extinction in $G_{\rm RP}$ band (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 84th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval.

ebpminrp_gspphot : Reddening $E(G_{\rm BP}-G_{\rm RP})$ from GSP-Phot Aeneas best library using BP/RP spectra (float, Magnitude[mag])

Reddening $E(G_{\rm BP}-G_{\rm RP})$ (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the median of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Note that while $E(G_{\rm BP}-G_{\rm RP})=A_{\rm BP}-A_{\rm RP}$, this was computed at the level of MCMC samples. Hence, this relation is not exactly true for the median values.

ebpminrp_gspphot_lower : Lower confidence level (16%) of reddening $E(G_{\rm BP}-G_{\rm RP})$ from GSP-Phot Aeneas best library using BP/RP spectra (float, Magnitude[mag])

Lower confidence level (16%) of reddening $E(G_{\rm BP}-G_{\rm RP})$ (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 16th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval. Note that while $E(G_{\rm BP}-G_{\rm RP})=A_{\rm BP}-A_{\rm RP}$, this was computed at the level of MCMC samples. Hence, this relation is not exactly true for the lower confidence levels.

ebpminrp_gspphot_upper : Upper confidence level (84%) of reddening $E(G_{\rm BP}-G_{\rm RP})$ from GSP-Phot Aeneas best library using BP/RP spectra (float, Magnitude[mag])

Upper confidence level (84%) of reddening $E(G_{\rm BP}-G_{\rm RP})$ (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 84th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval. Note that while $E(G_{\rm BP}-G_{\rm RP})=A_{\rm BP}-A_{\rm RP}$, this was computed at the level of MCMC samples. Hence, this relation is not exactly true for the upper confidence levels.

mg_gspphot : Absolute magnitude $M_{\rm G}$ from GSP-Phot Aeneas best library using BP/RP spectra (float, Magnitude[mag])

Absolute magnitude $M_{\rm G}$ (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the median of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value.

mg_gspphot_lower : Lower confidence level (16%) of absolute magnitude $M_{\rm G}$ from GSP-Phot Aeneas best library using BP/RP spectra (float, Magnitude[mag])

Lower confidence level (16%) of absolute magnitude $M_{\rm G}$ (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 16th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval.

mg_gspphot_upper : Upper confidence level (84%) of absolute magnitude $M_{\rm G}$ from GSP-Phot Aeneas best library using BP/RP spectra (float, Magnitude[mag])

Upper confidence level (84%) of absolute magnitude $M_{\rm G}$ (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 84th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval.

Stellar radius (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the median of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value.

radius_gspphot_lower : Lower confidence level (16%) of radius from GSP-Phot Aeneas best library using BP/RP spectra (float, Length & Distance[Solar Radius])

Lower confidence level (16%) of stellar radius (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 16th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval.

radius_gspphot_upper : Upper confidence level (84%) of radius from GSP-Phot Aeneas best library using BP/RP spectra (float, Length & Distance[Solar Radius])

Upper confidence level (84%) of stellar radius (assuming source is a single star) inferred by GSP-Phot Aeneas from BP/RP spectra, apparent G magnitude and parallax. This is the 84th percentile of the MCMC samples. Taken from best library that achieves the highest goodness-of-fit value. Lower and upper levels include 68% confidence interval.

logposterior_gspphot : Goodness-of-fit score (mean log-posterior of MCMC) of GSP-Phot Aeneas MCMC best library (float)

Goodness-of-fit score defined as the mean log-posterior of all MCMC samples of GSP-Phot Aeneas MCMC for best library. The higher the goodness-of-fit score, the better the fit. Values are usually negative. NB: This is not a Bayesian evidence!

mcmcaccept_gspphot : MCMC acceptance rate of GSP-Phot Aeneas MCMC best library (float)

MCMC acceptance rate of GSP-Phot Aeneas MCMC best library. This is computed from all MCMC samples (before thinning the chain to 2000 or 100 samples).

libname_gspphot : Name of library that achieves the highest mean log-posterior in MCMC samples and was used to derive GSP-Phot parameters in this table (string)

Name of library of synthetic stellar spectra (one of A, MARCS, OB, PHOENIX) for which GSP-Phot achieves the highest goodness-of-fit score (i.e. the highest mean log-posterior in its MCMC samples), referred to as “best library”. This is the library used to derive GSP-Phot parameters in this table (astrophysical_parameters). For more information on the synthetic libraries see Section 11.2.3.

teff_gspspec : Effective temperature from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Temperature[K])

Median value of the effective temperature (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) from RVS spectra and Monte Carlo realisations.

teff_gspspec_lower : 16th percentile of effective temperature from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Temperature[K])

Lower confidence level (16%) of the median effective temperature (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) from RVS spectra. Lower and upper levels include 68% confidence interval.

teff_gspspec_upper : 84th percentile of effective temperature from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Temperature[K])

Upper confidence level (84%) of the median effective temperature (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) from RVS spectra. Lower and upper levels include 68% confidence interval.

logg_gspspec : Logarithm of the stellar surface gravity from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, GravitySurface[log cgs])

Median value of logarithm of the stellar surface gravity (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) from RVS spectra.

logg_gspspec_lower : 16th percentile of the logarithm of the stellar surface gravity from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, GravitySurface[log cgs])

Lower confidence level (16%) of the median value of logarithm of the stellar surface gravity (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) from RVS spectra. Lower and upper levels include 68% confidence interval.

logg_gspspec_upper : 84th percentile of the logarithm of the stellar surface gravity from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, GravitySurface[log cgs])

Upper confidence level (84%) of the median value of logarithm of the stellar surface gravity (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) from RVS spectra. Lower and upper levels include 68% confidence interval.

mh_gspspec : Global metallicity [M/H] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Median global metallicity (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) from RVS spectra.

mh_gspspec_lower : 16th percentile of global metallicity [M/H] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Lower confidence level (16%) of the median global metallicity (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) from RVS spectra. Lower and upper levels include 68% confidence interval.

mh_gspspec_upper : 84th percentile of global metallicity [M/H] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Upper confidence level (84%) of the median global metallicity (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) from RVS spectra. Lower and upper levels include 68% confidence interval.

alphafe_gspspec : Abundance of alpha-elements [alpha/Fe] with respect to iron from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Median abundance of alpha-elements with respect to iron (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) from RVS spectra. The considered alpha elements are: O, Ne, Mg, Si, S, Ar, Ca, Ti.

alphafe_gspspec_lower : 16th percentile of the abundance of alpha-elements [alpha/Fe] with respect to iron from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Lower confidence level (16%) of the median abundance of alpha-elements with respect to iron (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) from RVS spectra. Lower and upper levels include 68% confidence interval.

alphafe_gspspec_upper : 84th percentile of the abundance of alpha-elements [alpha/Fe] with respect to iron from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Upper confidence level (84%) of the median abundance of alpha-elements with respect to iron (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) from RVS spectra. Lower and upper levels include 68% confidence interval.

fem_gspspec : Abundance of neutral iron [Fe/M] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations, applied to the individual N lines of the element, given in fem_gspspec_nlines (float, Abundances[dex])

Median abundance of neutral iron (assuming source is a single star) from RVS spectra and Monte Carlo realisations derived using MatisseGauguin (Recio-Blanco and et al. 2022) atmospheric parameters and the Gauguin algorithm, applied to the individual N lines of the element, where the number of lines is given in fem_gspspec_nlines. The neutral iron abundance [Fe/H] is obtained by [Fe/H]=[Fe/M]+[M/H].

fem_gspspec_lower : 16th percentile of the abundance of neutral iron [Fe/M] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Lower confidence level (16%) of the median abundance of neutral iron (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval. The neutral iron abundance [Fe/H] is obtained by [Fe/H]=[Fe/M]+[M/H].

fem_gspspec_upper : 84th percentile of the abundance of neutral iron [Fe/M] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Upper confidence level (84%) of the median abundance of neutral iron (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval. The neutral iron abundance [Fe/H] is obtained by [Fe/H]=[Fe/M]+[M/H].

fem_gspspec_nlines : Number of lines used for [Fe/M] abundance estimation (int)

Number of lines used to compute the [Fe/M] abundance. Lines with interquartile difference (84th quantile value - 16th quantile value) in the Monte Carlo line abundance distribution higher than 0.5 dex have been excluded.

fem_gspspec_linescatter : Uncertainty estimation of [Fe/M] abundance using N lines of the element, given in fem_gspspec_nlines (float, Abundances[dex])

Standard deviation of the individual N lines (fem_gspspec_nlines) abundance results.

sife_gspspec : Abundance of silicon [Si/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations, applied to the individual N lines of the element, given in sife_gspspec_nlines (float, Abundances[dex])

Median abundance of silicon (assuming source is a single star) from RVS spectra and Monte Carlo realisations derived using MatisseGauguin (Recio-Blanco and et al. 2022) atmospheric parameters and the Gauguin algorithm, applied to the individual N lines of the element, where the number of lines is given in sife_gspspec_nlines.

sife_gspspec_lower : 16th percentile of the abundance of silicon [Si/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Lower confidence level (16%) of the median abundance of silicon (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

sife_gspspec_upper : 84th percentile of the abundance of silicon [Si/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Upper confidence level (84%) of the median abundance of silicon (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

sife_gspspec_nlines : Number of lines used for [Si/Fe] abundance estimation (int)

Number of lines used to compute the [Si/Fe] abundance. Lines with interquartile difference (84th quantile value - 16th quantile value) in the Monte Carlo line abundance distribution higher than 0.5 dex have been excluded.

sife_gspspec_linescatter : Uncertainty estimation of [Si/Fe] abundance using N lines of the element, given in sife_gspspec_nlines (float, Abundances[dex])

Standard deviation of the individual N lines (sife_gspspec_nlines) abundance results.

cafe_gspspec : Abundance of calcium [Ca/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations, applied to the individual N lines of the element, given in cafe_gspspec_nlines (float, Abundances[dex])

Median abundance of calcium (assuming source is a single star) from RVS spectra and Monte Carlo realisations derived using MatisseGauguin (Recio-Blanco and et al. 2022) atmospheric parameters and the Gauguin algorithm, applied to the individual N lines of the element, where the number of lines is given in cafe_gspspec_nlines.

cafe_gspspec_lower : 16th percentile of the abundance of calcium [Ca/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Lower confidence level (16%) of the median abundance of calcium (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

cafe_gspspec_upper : 84th percentile of the abundance of calcium [Ca/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Upper confidence level (84%) of the median abundance of calcium (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

cafe_gspspec_nlines : Number of lines used for [Ca/Fe] abundance estimation (int)

Number of lines used to compute the [Ca/Fe] abundance. Lines with interquartile difference (84th quantile value - 16th quantile value) in the Monte Carlo line abundance distribution higher than 0.5 dex have been excluded.

cafe_gspspec_linescatter : Uncertainty estimation of [Ca/Fe] abundance using N lines of the element, given in cafe_gspspec_nlines (float, Abundances[dex])

Standard deviation of the individual N lines (cafe_gspspec_nlines) abundance results.

tife_gspspec : Abundance of titanium [Ti/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations, applied to the individual N lines of the element, given in tife_gspspec_nlines (float, Abundances[dex])

Median abundance of titanium (assuming source is a single star) from RVS spectra and Monte Carlo realisations derived using MatisseGauguin (Recio-Blanco and et al. 2022) atmospheric parameters and the Gauguin algorithm, applied to the individual N lines of the element, where the number of lines is given in tife_gspspec_nlines.

tife_gspspec_lower : 16th percentile of the abundance of titanium [Ti/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Lower confidence level (16%) of the median abundance of titanium (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

tife_gspspec_upper : 84th percentile of the abundance of titanium [Ti/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Upper confidence level (84%) of the median abundance of titanium (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

tife_gspspec_nlines : Number of lines used for [Ti/Fe] abundance estimation (int)

Number of lines used to compute the [Ti/Fe] abundance. Lines with interquartile difference (84th quantile value - 16th quantile value) in the Monte Carlo line abundance distribution higher than 0.5 dex have been excluded.

tife_gspspec_linescatter : Uncertainty estimation of [Ti/Fe] abundance using N lines of the element, given in tife_gspspec_nlines (float, Abundances[dex])

Standard deviation of the individual N lines (tife_gspspec_nlines) abundance results.

mgfe_gspspec : Abundance of magnesium [Mg/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations, applied to the individual N lines of the element, given in mgfe_gspspec_nlines (float, Abundances[dex])

Median abundance of magnesium (assuming source is a single star) from RVS spectra and Monte Carlo realisations derived using MatisseGauguin (Recio-Blanco and et al. 2022) atmospheric parameters and the Gauguin algorithm, applied to the individual N lines of the element, where the number of lines is given in mgfe_gspspec_nlines.

mgfe_gspspec_lower : 16th percentile of the abundance of magnesium [Mg/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Lower confidence level (16%) of the median abundance of magnesium (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

mgfe_gspspec_upper : 84th percentile of the abundance of magnesium [Mg/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Upper confidence level (84%) of the median abundance of magnesium (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

mgfe_gspspec_nlines : Number of lines used for [Mg/Fe] abundance estimation (int)

Number of lines used to compute the [Mg/Fe] abundance. Lines with interquartile difference (84th quantile value - 16th quantile value) in the Monte Carlo line abundance distribution higher than 0.5 dex have been excluded.

mgfe_gspspec_linescatter : Uncertainty estimation of [Mg/Fe] abundance using N lines of the element, given in mgfe_gspspec_nlines (float, Abundances[dex])

Standard deviation of the individual N lines (mgfe_gspspec_nlines) abundance results.

ndfe_gspspec : Abundance of neodymium [Nd/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations, applied to the individual N lines of the element, given in ndfe_gspspec_nlines (float, Abundances[dex])

Median abundance of neodymium (assuming source is a single star) from RVS spectra and Monte Carlo realisations derived using MatisseGauguin (Recio-Blanco and et al. 2022) atmospheric parameters and the Gauguin algorithm, applied to the individual N lines of the element, where the number of lines is given in ndfe_gspspec_nlines.

ndfe_gspspec_lower : 16th percentile of the abundance of neodymium [Nd/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Lower confidence level (16%) of the median abundance of neodymium (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

ndfe_gspspec_upper : 84th percentile of the abundance of neodymium [Nd/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Upper confidence level (84%) of the median abundance of neodymium (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

ndfe_gspspec_nlines : Number of lines used for [Nd/Fe] abundance estimation (int)

Number of lines used to compute the [Nd/Fe] abundance. Lines with interquartile difference (84th quantile value - 16th quantile value) in the Monte Carlo line abundance distribution higher than 0.5 dex have been excluded.

ndfe_gspspec_linescatter : Uncertainty estimation of [Nd/Fe] abundance using N lines of the element, given in ndfe_gspspec_nlines (float, Abundances[dex])

Standard deviation of the individual N lines (ndfe_gspspec_nlines) abundance results.

feiim_gspspec : Abundance of ionised iron [FeII/M] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations, applied to the individual N lines of the element, given in feiim_gspspec_nlines (float, Abundances[dex])

Median abundance of ionised iron (assuming source is a single star) from RVS spectra and Monte Carlo realisations derived using MatisseGauguin (Recio-Blanco and et al. 2022) atmospheric parameters and the Gauguin algorithm, applied to the individual N lines of the element, where the number of lines is given in feiim_gspspec_nlines. The ionised iron abundance [FeII/H] is obtained by [FeII/H]=[FeII/M]+[M/H].

feiim_gspspec_lower : 16th percentile of the abundance of ionised iron [FeII/M] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Lower confidence level (16%) of the median abundance of ionised iron (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval. The ionised iron abundance [FeII/H] is obtained by [FeII/H]=[FeII/M]+[M/H].

feiim_gspspec_upper : 84th percentile of the abundance of ionised iron [FeII/M] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Upper confidence level (84%) of the median abundance of ionised iron (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval. The ionised iron abundance [FeII/H] is obtained by [FeII/H]=[FeII/M]+[M/H].

feiim_gspspec_nlines : Number of lines used for [FeII/M] abundance estimation (int)

Number of lines used to compute the [FeII/M] abundance. Lines with interquartile difference (84th quantile value - 16th quantile value) in the Monte Carlo line abundance distribution higher than 0.5 dex have been excluded.

feiim_gspspec_linescatter : Uncertainty estimation of [FeII/M] abundance using N lines of the element, given in feiim_gspspec_nlines (float, Abundances[dex])

Standard deviation of the individual N lines (feiim_gspspec_nlines) abundance results.

sfe_gspspec : Abundance of sulphur [S/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations, applied to the individual N lines of the element, given in sfe_gspspec_nlines (float, Abundances[dex])

Median abundance of sulphur (assuming source is a single star) from RVS spectra and Monte Carlo realisations derived using MatisseGauguin (Recio-Blanco and et al. 2022) atmospheric parameters and the Gauguin algorithm, applied to the individual N lines of the element, where the number of lines is given in sfe_gspspec_nlines.

sfe_gspspec_lower : 16th percentile of the abundance of sulphur [S/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Lower confidence level (16%) of the median abundance of sulphur (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

sfe_gspspec_upper : 84th percentile of the abundance of sulphur [S/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Upper confidence level (84%) of the median abundance of sulphur (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

sfe_gspspec_nlines : Number of lines used for [S/Fe] abundance estimation (int)

Number of lines used to compute the [S/Fe] abundance. Lines with interquartile difference (84th quantile value - 16th quantile value) in the Monte Carlo line abundance distribution higher than 0.5 dex have been excluded.

sfe_gspspec_linescatter : Uncertainty estimation of [S/Fe] abundance using N lines of the element, given in sfe_gspspec_nlines (float, Abundances[dex])

Standard deviation of the individual N lines (sfe_gspspec_nlines) abundance results.

zrfe_gspspec : Abundance of zirconium [Zr/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations, applied to the individual N lines of the element, given in zrfe_gspspec_nlines (float, Abundances[dex])

Median abundance of zirconium (assuming source is a single star) from RVS spectra and Monte Carlo realisations derived using MatisseGauguin (Recio-Blanco and et al. 2022) atmospheric parameters and the Gauguin algorithm, applied to the individual N lines of the element, where the number of lines is given in zrfe_gspspec_nlines.

zrfe_gspspec_lower : 16th percentile of the abundance of zirconium [Zr/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Lower confidence level (16%) of the median abundance of zirconium (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

zrfe_gspspec_upper : 84th percentile of the abundance of zirconium [Zr/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Upper confidence level (84%) of the median abundance of zirconium (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

zrfe_gspspec_nlines : Number of lines used for [Zr/Fe] abundance estimation (int)

Number of lines used to compute the [Zr/Fe] abundance. Lines with interquartile difference (84th quantile value - 16th quantile value) in the Monte Carlo line abundance distribution higher than 0.5 dex have been excluded.

zrfe_gspspec_linescatter : Uncertainty estimation of [Zr/Fe] abundance using N lines of the element, given in zrfe_gspspec_nlines (float, Abundances[dex])

Standard deviation of the individual N lines (zrfe_gspspec_nlines) abundance results.

nfe_gspspec : Abundance of nitrogen [N/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations, applied to the individual N lines of the element, given in nfe_gspspec_nlines (float, Abundances[dex])

Median abundance of nitrogen (assuming source is a single star) from RVS spectra and Monte Carlo realisations derived using MatisseGauguin (Recio-Blanco and et al. 2022) atmospheric parameters and the Gauguin algorithm, applied to the individual N lines of the element, where the number of lines is given in nfe_gspspec_nlines.

nfe_gspspec_lower : 16th percentile of the abundance of nitrogen [N/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Lower confidence level (16%) of the median abundance of nitrogen (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

nfe_gspspec_upper : 84th percentile of the abundance of nitrogen [N/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Upper confidence level (84%) of the median abundance of nitrogen (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

nfe_gspspec_nlines : Number of lines used for [N/Fe] abundance estimation (int)

Number of lines used to compute the [N/Fe] abundance. Lines with interquartile difference (84th quantile value - 16th quantile value) in the Monte Carlo line abundance distribution higher than 0.5 dex have been excluded.

nfe_gspspec_linescatter : Uncertainty estimation of [N/Fe] abundance using N lines of the element, given in nfe_gspspec_nlines (float, Abundances[dex])

Standard deviation of the individual N lines (nfe_gspspec_nlines) abundance results.

crfe_gspspec : Abundance of chromium [Cr/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations, applied to the individual N lines of the element, given in crfe_gspspec_nlines (float, Abundances[dex])

Median abundance of chromium (assuming source is a single star) from RVS spectra and Monte Carlo realisations derived using MatisseGauguin (Recio-Blanco and et al. 2022) atmospheric parameters and the Gauguin algorithm, applied to the individual N lines of the element, where the number of lines is given in crfe_gspspec_nlines.

crfe_gspspec_lower : 16th percentile of the abundance of chromium [Cr/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Lower confidence level (16%) of the median abundance of chromium (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

crfe_gspspec_upper : 84th percentile of the abundance of chromium [Cr/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Upper confidence level (84%) of the median abundance of chromium (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

crfe_gspspec_nlines : Number of lines used for [Cr/Fe] abundance estimation (int)

Number of lines used to compute the [Cr/Fe] abundance. Lines with interquartile difference (84th quantile value - 16th quantile value) in the Monte Carlo line abundance distribution higher than 0.5 dex have been excluded.

crfe_gspspec_linescatter : Uncertainty estimation of [Cr/Fe] abundance using N lines of the element, given in crfe_gspspec_nlines (float, Abundances[dex])

Standard deviation of the individual N lines (crfe_gspspec_nlines) abundance results.

cefe_gspspec : Abundance of cerium [Ce/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations, applied to the individual N lines of the element, given in cefe_gspspec_nlines (float, Abundances[dex])

Median abundance of cerium (assuming source is a single star) from RVS spectra and Monte Carlo realisations derived using MatisseGauguin (Recio-Blanco and et al. 2022) atmospheric parameters and the Gauguin algorithm, applied to the individual N lines of the element, where the number of lines is given in cefe_gspspec_nlines.

cefe_gspspec_lower : 16th percentile of the abundance of cerium [Ce/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Lower confidence level (16%) of the median abundance of cerium (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

cefe_gspspec_upper : 84th percentile of the abundance of cerium [Ce/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Upper confidence level (84%) of the median abundance of cerium (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

cefe_gspspec_nlines : Number of lines used for [Ce/Fe] abundance estimation (int)

Number of lines used to compute the [Ce/Fe] abundance. Lines with interquartile difference (84th quantile value - 16th quantile value) in the Monte Carlo line abundance distribution higher than 0.5 dex have been excluded.

cefe_gspspec_linescatter : Uncertainty estimation of [Ce/Fe] abundance using N lines of the element, given in cefe_gspspec_nlines (float, Abundances[dex])

Standard deviation of the individual N lines (cefe_gspspec_nlines) abundance results.

nife_gspspec : Abundance of nickel [Ni/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations, applied to the individual N lines of the element, given in nife_gspspec_nlines (float, Abundances[dex])

Median abundance of nickel (assuming source is a single star) from RVS spectra and Monte Carlo realisations derived using MatisseGauguin (Recio-Blanco and et al. 2022) atmospheric parameters and the Gauguin algorithm, applied to the individual N lines of the element, where the number of lines is given in nife_gspspec_nlines.

nife_gspspec_lower : 16th percentile of the abundance of nickel [Ni/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Lower confidence level (16%) of the median abundance of nickel (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

nife_gspspec_upper : 84th percentile of the abundance of nickel [Ni/Fe] from GSP-Spec MatisseGauguin using RVS spectra and Monte Carlo realisations (float, Abundances[dex])

Upper confidence level (84%) of the median abundance of nickel (assuming source is a single star) inferred by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) using RVS spectra. Lower and upper levels include 68% confidence interval.

nife_gspspec_nlines : Number of lines used for [Ni/Fe] abundance estimation (int)

Number of lines used to compute the [Ni/Fe] abundance. Lines with interquartile difference (84th quantile value - 16th quantile value) in the Monte Carlo line abundance distribution higher than 0.5 dex have been excluded.

nife_gspspec_linescatter : Uncertainty estimation of [Ni/Fe] abundance using N lines of the element, given in nife_gspspec_nlines (float, Abundances[dex])

Standard deviation of the individual N lines (nife_gspspec_nlines) abundance results.

cn0ew_gspspec : Equivalent witdh of cyanogen absorption line, derived from RVS spectra (float, Length & Distance[nm])

Equivalent width of the residual feature (computed as the observed RVS spectrum divided by a synthetic one with the MatisseGauguin parameters) around the cyanogen line at 862.9 nm.

cn0ew_gspspec_uncertainty : Uncertainty of equivalent witdh of cyanogen absorption line, derived from RVS spectra (float, Length & Distance[nm])

Interquartile difference (84th quantile value - 16th quantile value) in the Monte Carlo distribution of cn0ew_gspspec, derived from RVS spectra.

cn0_gspspec_centralline : Central wavelength of cyanogen line, derived from RVS spectra using DIB algorithm (float, Length & Distance[nm])

Central wavelength of the Gaussian fit applied to the residual feature (computed as the observed RVS spectrum divided by a synthetic one with the MatisseGauguin parameters) around the cyanogen line at 862.9 nm.

cn0_gspspec_width : Width of cyoanogen line, derived from RVS spectra using DIB algorithm (float, Length & Distance[nm])

Width of the Gaussian fit applied to the residual feature (computed as the observed RVS spectrum divided by a synthetic one with the MatisseGauguin parameters) around the cyanogen line at 862.9 nm.

dib_gspspec_lambda : DIB central wavelength from GSP-Spec MatisseGauguin using RVS spectra (float, Length & Distance[nm])

Central wavelength of the DIB feature in the RVS spectrum derived by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022).

dib_gspspec_lambda_uncertainty : Uncertainty on DIB central wavelength from GSP-Spec MatisseGauguin using RVS spectra (float, Length & Distance[nm])

Uncertainty on central wavelength of the DIB feature in the RVS spectrum derived by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022).

dibew_gspspec : Equivalent width of the DIB from GSP-Spec MatisseGauguin using RVS spectra (float, Length & Distance[Å])

Equivalent width of the DIB feature in the RVS spectrum derived by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022).

dibew_gspspec_uncertainty : Global uncertainty on DIB equivalent width value using DIB algorithm (float, Length & Distance[Å])

Global uncertainty on equivalent width of the DIB feature in the RVS spectrum derived by GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022).

dibewnoise_gspspec_uncertainty : Uncertainty on DIB equivalent width value occuring from noise part (float, Length & Distance[Å])

Uncertainty on DIB equivalent width value based on the noise level.

dibp0_gspspec : Depth ($p_{0}$ parameter) of the DIB derived from a Gaussian model fit (float)

Depth ($p_{0}$ parameter) of the DIB defined from a Gaussian model fit. The flux is modelled as $p_{0}\times\exp\left(-\frac{(x-p_{1})^{2}}{2p_{2}^{2}}\right),$ where $p_{0}$ and $p_{2}$ are the depth and width of the DIB profile, $p_{1}$ is the central wavelength and $x$ is the spectral wavelength, cf. Zhao et al. (2021).

dibp2_gspspec : Width ($p_{2}$ parameter) of the DIB derived from a Gaussian model fit (float, Length & Distance[Å])

Width ($p_{2}$ parameter) of the DIB defined from a Gaussian model fit. The flux is modelled as $p_{0}\times\exp\left(-\frac{(x-p_{1})^{2}}{2p_{2}^{2}}\right),$ where $p_{0}$ and $p_{2}$ are the depth and width of the DIB profile, $p_{1}$ is the central wavelength and $x$ is the spectral wavelength, cf. Zhao et al. (2021).

dibp2_gspspec_uncertainty : Uncertainty on the dibp2_gspspec parameter (float, Length & Distance[Å])

Uncertainty on the $p_{2}$ parameter from the Gaussian fitting, given in dibp2_gspspec.

dibqf_gspspec : Quality flag of the DIB computation (int)

Quality flag on DIB computation: QF=$-1$ means that there is not a preliminary detection where sources are only measured if the detection threshold is above the 3-sigma level, QF=$-2$ means outside the considered temperature range, i.e., $T_{\rm eff}<3500$ K, or flux values are NaN in the DIB wavelength range between 860.5 and 864.0 nm.

flags_gspspec : Catalogue flags for GSP-Spec MatisseGauguin (string)

Definitions of each character in the GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022) quality flag chain. In this chain, value ‘0’ is the best, and ‘9’ is the worst. Flag names are split in three categories: parameter flags (green), individual abundance flags (blue) and equivalent width flags (maroon): In the following, the term ’gof’ refers to the chi-square value beteween the input spectrum fluxes array and the solution spectrum fluxes array.

 Chain character Considered Possible number - name quality aspect adopted values 1 vbroadT vbroad induced bias in $T_{\rm eff}$ 0,1,2,9 2 vbroadG vbroad induced bias in $\log g$ 0,1,2,9 3 vbroadM vbroad induced bias in [M/H] 0,1,2,9 4 vradT vrad induced bias in $T_{\rm eff}$ 0,1,2,9 5 vradG vrad induced bias in $\log g$ 0,1,2,9 6 vradM vrad induced bias in [M/H] 0,1,2,9 7 fluxNoise flux noise uncertainties 0,1,2,3,4,5,9 8 extrapol extrapolation 0,1,2,3,4,9 9 neg_flux negative flux pixels 0,1,9 10 nanFlux NaN flux pixels 0,9 11 emission emission line 0,9 12 nullFluxErr null uncertainties 0,9 13 KMgiantPar KM-type giant stars 0,1,2 14 NUpLim nitrogen abundance upper limit 0,1,2,9 15 NUncer nitrogen abundance uncertainty quality 0,1,2,9 16 MgUpLim magnesium abundance upper limit 0,1,2,9 17 MgUncer magnesium abundance uncertainty quality 0,1,2,9 18 SiUpLim silicon abundance upper limit 0,1,2,9 19 SiUncer silicon abundance uncertainty quality 0,1,2,9 20 SUpLim sulphur abundance upper limit 0,1,2,9 21 SUncer sulphur abundance uncertainty quality 0,1,2,9 22 CaUpLim calcium abundance upper limit 0,1,2,9 23 CaUncer calcium abundance uncertainty quality 0,1,2,9 24 TiUpLim titanium abundance upper limit 0,1,2,9 25 TiUncer titanium abundance uncertainty quality 0,1,2,9 26 CrUpLim chromium abundance upper limit 0,1,2,9 27 CrUncer chromium abundance uncertainty quality 0,1,2,9 28 FeUpLim neutral iron abundance upper limit 0,1,2,9 29 FeUncer neutral iron abundance uncertainty quality 0,1,2,9 30 FeIIUpLim ionised iron abundance upper limit 0,1,2,9 31 FeIIUncer ionised iron abundance uncertainty quality 0,1,2,9 32 NiUpLim nickel abundance upper limit 0,1,2,9 33 NiUncer nickel abundance uncertainty quality 0,1,2,9 34 ZrUpLim zirconium abundance upper limit 0,1,2,9 35 ZrUncer zirconium abundance uncertainty quality 0,1,2,9 36 CeUpLim cerium abundance upper limit 0,1,2,9 37 CeUncer cerium abundance uncertainty quality 0,1,2,9 38 NdUpLim neodymium abundance upper limit 0,1,2,9 39 NdUncer neodymium abundance uncertainty quality 0,1,2,9 40 DeltaCNq cyanogen differential equivalent width quality 0,9 41 DIBq DIB quality flag 0,1,2,3,4,5,9

Definition of parameter flags considering potential biases due to rotational velocity and macroturbulence:

 Flag name Condition Flag value vbroadT $\Delta$$T_{\rm eff}$$>$2000 K Filter all Flag $=$ 9 500$<\Delta$$T_{\rm eff}$$\leq$2000 K Flag $=$ 2 250$<\Delta$$T_{\rm eff}$$\leq$500 K Flag $=$ 1 $\Delta$$T_{\rm eff}$$\leq$250 K Flag $=$ 0 vbroadG $\Delta$$\log g$$>$2 dex Filter all except $T_{\rm eff}$ and DIB if $T_{\rm eff}$$>$7000 K Flag $=$ 9 1$<\Delta$$\log g$$\leq$2 dex Flag $=$ 2 0.5$<\Delta$$\log g$$\leq$1 dex Flag $=$ 1 $\Delta$$\log g$$\leq$0.5 dex Flag $=$ 0 vbroadM $\Delta$[M/H]$>$2 dex Filter [M/H] and [X/Fe] Flag $=$ 9 0.5$<\Delta$[M/H]$\leq$2 dex Flag $=$ 2 0.25$<\Delta$[M/H]$\leq$0.5 dex Flag $=$ 1 $\Delta$[M/H]$\leq$0.25 dex Flag $=$ 0

Definition of parameter flags considering potential biases due to uncertainties in the radial velocity shift correction:

 Flag name Condition Flag value vradT $\Delta$$T_{\rm eff}$$>$2000 K Filter all Flag $=$ 9 500$<\Delta$$T_{\rm eff}$$\leq$2000 K Flag $=$ 2 250$<\Delta$$T_{\rm eff}$$\leq$500 K Flag $=$ 1 $\Delta$$T_{\rm eff}$$\leq$250 K Flag $=$ 0 vradG $\Delta$$\log g$$>$2 dex Filter all except $T_{\rm eff}$ and DIB if $T_{\rm eff}$$>$7000 K Flag $=$ 9 1$<\Delta$$\log g$$\leq$2 dex Flag $=$ 2 0.5$<\Delta$$\log g$$\leq$1 dex Flag $=$ 1 $\Delta$$\log g$$\leq$0.5 dex Flag $=$ 0 vradM $\Delta$[M/H]$>$2 dex Filter [M/H] and [X/Fe] Flag $=$ 9 0.5$<\Delta$[M/H]$\leq$2 dex Flag $=$ 2 0.25$<\Delta$[M/H]$\leq$0.5 dex Flag $=$ 1 $\Delta$[M/H]$\leq$0.25 dex Flag $=$ 0

Definition of parameter flags considering potential biases due to uncertainties in the RVS flux:

 Flag name Condition Flag value fluxNoise $\sigma$$T_{\rm eff}$$>$2000 K or Filter all $\sigma$$\log g$$>$2 dex Flag $=$ 9 $\sigma$$T_{\rm eff}$$\leq$2000 K and $\sigma$$\log g$$\leq$2 dex and Filter [M/H], [X/Fe] $\sigma$[M/H]$>$2 dex Flag $=$ 5 $\sigma$$T_{\rm eff}$$\leq$2000 K and $\sigma$$\log g$$\leq$2 dex and Filter [$\alpha$/Fe], [X/Fe] $\sigma$[M/H]$\leq$2 dex and Flag $=$ 4 $\sigma$[$\alpha$/Fe]$>$0.8 dex 500$<\sigma$$T_{\rm eff}$$\leq$2000 K and Flag $=$ 3 1$<\sigma$$\log g$$\leq$2 dex and 0.5$<\sigma$[M/H]$\leq$2 dex and 0.2$<\sigma$[$\alpha$/Fe]$\leq$0.8 dex 250$<\sigma$$T_{\rm eff}$$\leq$500 K and Flag $=$ 2 0.5$<\sigma$$\log g$$\leq$1 dex and 0.25$<\sigma$[M/H]$\leq$0.5 dex and 0.1$<\sigma$[$\alpha$/Fe]$\leq$0.2 dex 100$<\sigma$$T_{\rm eff}$$\leq$250 K and Flag $=$ 1 0.2$<\sigma$$\log g$$\leq$0.5 dex and 0.1$<\sigma$[M/H]$\leq$0.25 dex and 0.05$<\sigma$[$\alpha$/Fe]$\leq$0.1 dex $\sigma$$T_{\rm eff}$$\leq$100 K and Flag $=$ 0 $\sigma$$\log g$$\leq$0.2 dex and $\sigma$[M/H]$\leq$0.1 dex and $\sigma$[$\alpha$/Fe]$\leq$0.05 dex

Definition of parameter flags considering potential biases due to extrapolated parameters:

 Flag name Condition Flag value extrapol gof$=$NaN and Filter all ($T_{\rm eff}$$>$9000 K or $T_{\rm eff}$$<$2500 K or except DIB if $T_{\rm eff}$$>$7000 K $\log g$$>$6 dex or $\log g$$<-1$ dex ) Flag $=$ 9 gof$=$NaN and 2500$\leq$$T_{\rm eff}$$\leq$9000 K and Filter [M/H],[X/Fe] $-1\leq$$\log g$$\leq$6 dex and Flag $=$ 4 ([M/H]$<-6$ dex or [M/H]$>$1.5 dex ) gof$=$NaN and 2500$\leq$$T_{\rm eff}$$<$9000 K and $-1\leq$$\log g$$\leq$6 dex and Filter [X/Fe] $-6\leq$[M/H]$\leq$1.5 dex and Flag $=$ 3 [$\alpha$/Fe] out from standard by $\pm$ 0.8 gof$=$NaN and Flag $=$ 2 2500$\leq$$T_{\rm eff}$$\leq$9000 K and $-1\leq$$\log g$$\leq$6 dex and $-6\leq$[M/H]$\leq$1.5 dex and [$\alpha$/Fe] within $\pm$ 0.8 from standard gof$\neq$NaN and Flag $=$ 1 ($T_{\rm eff}$$\geq$7625 or $T_{\rm eff}$$\leq$3500 K or $\log g$$\geq$4.75 or $\log g$$\leq$0.25 dex or [M/H]$\leq-3$ or [M/H]$\geq$0.75 dex or [$\alpha$/Fe] out from standard by $\pm$ 0.35) gof$\neq$NaN and Flag $=$ 0 3500$<$$T_{\rm eff}$$<$7625 K and 0.25$<$$\log g$$<$4.75 dex and $-3<$[M/H]$<$0.75 dex and [$\alpha$/Fe] within $\pm$ 0.35 from standard

Definition of parameter flags considering RVS flux problems or emission line probability:

 Flag name Condition Flag value nanFlux Flux$=$NaN Filter all except DIB if $T_{\rm eff}$$>$7000 K Flag $=$ 9 no Nans in Flux Flag $=$ 0 emission CU6_is_emission$=$True Filter all except DIB if $T_{\rm eff}$$>$7000 K Flag $=$ 9 CU6_is_emission$=$False Flag $=$ 0 neg_flux $>$ 2 pixels with flux$<$0 Filter all except DIB if $T_{\rm eff}$$>$7000 K Flag $=$ 9 1 or 2 pixels with flux$<$0 Flag $=$ 1 flux$>$0 Flag $=$ 0 nullFluxErr $\sigma$$T_{\rm eff}$$=$0 K or $\sigma$$\log g$$=$0 dex or Filter all $\sigma$[M/H]$=$0 dex or Flag $=$ 9 $\sigma$[$\alpha$/Fe]$=$0 dex no null uncertainties for all pixel flux Flag $=$ 0

Definition of parameter flags considering problems in the parameterisation of KM-type giants. $F_{\rm min}$ is the minumum flux value in the corresponding RVS spectrum:

 Flag name Condition Flag value KMgiantPar $T_{\rm eff}$$<$4000 K and $\log g$$<$3.5 and Filter [$\alpha$/Fe] (gof$>-3.0$ or $F_{\rm min}$$>$0.22) Flag $=$ 2 $T_{\rm eff}$$<$4000 K and $\log g$$<$3.5 and ($-3.4<$gof$<-3.0$ or Filter [$\alpha$/Fe] (gof$>-3.0$ and $F_{\rm min}$$<$0.22) or Flag $=$ 1 (gof$<-3.4$ and $F_{\rm min}$$>$0.22)) ($T_{\rm eff}$$<$4000 K and $\log g$$<$3.5 and Flag $=$ 0 (gof$<-3.4$ or $F_{\rm min}$$<$0.22)) or $T_{\rm eff}$$<$4000 or $\log g$$>$3.5

Definition of individual abundance upper limit flags. Xfe_gspspec_upper is the upper confidence value of the abundance, corresponding to the 84th quantile of the Monte Carlo distribution. $\sigma$[X/Fe] is the 84th$-$16th interquantile abundance uncertainty. XfeUpperLimit is the mean value of the abundance upper limit for the considered lines of element X in the spectrum, depending on the mean signal-to-noise ratio (SNR) in the line pixels and on the stellar parameters. X_MAD_UpperLimit is the median absolute deviation of the upper limit in the line pixels; the coefficients c1 to c8 are reported in a table further below:

 Flag name Condition Flag value XUpLim vbroadT$\geq$2 or vbroadG$\geq$2 or vbroadM$\geq$2 or $\sigma$[X/Fe]$=$0 or (Xfe_gspspec_upper$-$Xfe_gspspec)$=$ 0 or $T_{\rm eff}$$\leq$c1 or $T_{\rm eff}$$\geq$c2 or $\log g$$\leq$c3 or $\log g$$\geq$c4 or Filter [X/Fe] ((2$-$ XfeUpperLimit)/$\sigma$[X/Fe])$\leq$c5 or Flag $=$ 9 (SNR$\leq$c6 and gof$\geq$c7) or (Xfe_gspspec+[M/H])$\leq$c8 vbroadT$<$2 and vbroadG$<$2 and vbroadM$<$2 and Flag $=$ 2 $\sigma$[X/Fe]$\neq$0 and (Xfe_gspspec_upper$-$ Xfe_gspspec)$\neq$ 0 and c1$<$$T_{\rm eff}$$<$c2 and c3$<$$\log g$$<$c4 and ((2$-$ XfeUpperLimit)/$\sigma$[X/Fe])$>$c5 and (SNR$>$c6 or (SNR$\leq$c6 and gof$<$c7)) and (Xfe_gspspec+[M/H])$<$c8 and ((Xfe_gspspec$-$ XfeUpperLimit)/(1.48$\cdot$X_MAD_UpperLimit))$<$1.5 vbroadT$<$2 and vbroadG$<$2 and vbroadM$<$2 and Flag $=$ 1 $\sigma$[X/Fe]$\neq$0 and (Xfe_gspspec_upper$-$ Xfe_gspspec)$\neq$ 0 and c1$<$$T_{\rm eff}$$<$c2 and c3$<$$\log g$$<$c4 and ((2$-$ XfeUpperLimit)/$\sigma$[X/Fe])$>$c5 and (SNR$>$c6 or (SNR$\leq$c6 and gof$<$c7)) and (Xfe_gspspec+[M/H])$<$c8 and 1.5$\leq$((Xfe_gspspec$-$ XfeUpperLimit)/(1.48$\cdot$X_MAD_UpperLimit))$<$2.5 vbroadT$<$2 and vbroadG$<$2 and vbroadM$<$2 and Flag $=$ 0 $\sigma$[X/Fe]$\neq$0 and (Xfe_gspspec_upper$-$ Xfe_gspspec)$\neq$ 0 and c1$<$$T_{\rm eff}$$<$c2 and c3$<$$\log g$$<$c4 and ((2$-$ XfeUpperLimit)/$\sigma$[X/Fe])$>$c5 and (SNR$>$c6 or (SNR$\leq$c6 and gof$<$c7)) and (Xfe_gspspec+[M/H])$<$c8 and ((Xfe_gspspec$-$ XfeUpperLimit)/(1.48$\cdot$X_MAD_UpperLimit))$\geq$2.5

Definition of individual abundance uncertainty flags. Xfe_gspspec_upper is the upper confidence value of the abundance, corresponding to the 84th quantile of the Monte Carlo distribution. $\sigma$[X/Fe] is the 84th$-$16th interquantile abundance uncertainty. XfeUpperLimit is the mean value of the abundance upper limit for the considered lines of element X in the spectrum, depending on the mean SNR in the line pixels and the stellar parameters. X_MAD_UpperLimit is the median absolute deviation of the upper limit in the line pixels; the coefficients c1 to c8 are reported in a table further below:

 Flag name Condition Flag value XUncer vbroadT$\geq$2 or vbroadG$\geq$2 or vbroadM$\geq$2 or $\sigma$[X/Fe]$=$0 or (Xfe_gspspec_upper$-$Xfe_gspspec)$=$ 0 or $T_{\rm eff}$$\leq$c1 or $T_{\rm eff}$$\geq$c2 or $\log g$$\leq$c3 or $\log g$$\geq$c4 or Filter [X/Fe] ((2$-$XfeUpperLimit)/$\sigma$[X/Fe])$\leq$c5 or Flag $=$ 9 (SNR$\leq$c6 and gof$\geq$c7) or (Xfe_gspspec+[M/H])$\leq$c8 vbroadT$<$2 and vbroadG$<$2 and vbroadM$<$2 and Flag $=$ 2 $\sigma$[X/Fe]$\neq$0 and (Xfe_gspspec_upper$-$Xfe_gspspec)$\neq$ 0 and c1$<$$T_{\rm eff}$$<$c2 and c3$<$$\log g$$<$c4 and c5$<$((2$-$XfeUpperLimit)/$\sigma$[X/Fe])$<$7 and (SNR$>$c6 or (SNR$\leq$c6 and gof$<$c7)) and (Xfe_gspspec+[M/H])$<$c8 vbroadT$<$2 and vbroadG$<$2 and vbroadM$<$2 and Flag $=$ 1 $\sigma$[X/Fe]$\neq$0 and (Xfe_gspspec_upper$-$Xfe_gspspec)$\neq$ 0 and c1$<$$T_{\rm eff}$$<$c2 and c3$<$$\log g$$<$c4 and 7$\leq$((2$-$XfeUpperLimit)/$\sigma$[X/Fe])$<$10 and (SNR$>$c6 or (SNR$\leq$c6 and gof$<$c7)) and (Xfe_gspspec+[M/H])$<$c8 vbroadT$<$2 and vbroadG$<$2 and vbroadM$<$2 and Flag $=$ 0 $\sigma$[X/Fe]$\neq$0 and (Xfe_gspspec_upper$-$Xfe_gspspec)$\neq$ 0 and c1$<$$T_{\rm eff}$$<$c2 and c3$<$$\log g$$<$c4 and ((2$-$XfeUpperLimit)/$\sigma$[X/Fe])$\geq$10 and (SNR$>$c6 or (SNR$\leq$c6 and gof$<$c7)) and (Xfe_gspspec+[M/H])$<$c8

Coefficients for individual chemical abundance filtering used in the previous two tables:

 Chemical abundance c1 c2 c3 c4 c5 c6 c7 c8 [N/Fe] 4200 8000 0.0 5.5 4.5 100 $-3.6$ 99 [Mg/Fe] 3500 8000 $-1.0$ 5.5 5.5 80 $-3.5$ 99 [Si/Fe] 4000 8000 $-1.0$ 5.5 6.0 110 $-3.8$ 99 [S/Fe] 5500 8000 3.0 5.5 5.0 120 $-3.7$ 99 [Ca/Fe] 3500 8000 $-1.0$ 5.5 10.0 60 $-3.2$ 99 [Ti/Fe] 4000 6500 $-1.0$ 5.5 6.0 110 $-3.65$ 99 [Cr/Fe] 3500 6000 $-1.0$ 5.5 6.0 1000 $-3.65$ 1.5 [Fe/M] 3500 8000 $-1.0$ 5.5 5.0 1000 $-3.4$ 1.5 [FeII/M] 5700 8000 3.5 5.5 5.0 70 $-3.5$ 1.5 [Ni/Fe] 4000 6500 $-1.0$ 5.5 6.0 100 $-3.6$ 1.5 [Zr/Fe] 3500 8000 $-1.0$ 5.5 1.0 100 $-3.4$ 99 [Ce/Fe] 3500 8000 $-1.0$ 5.5 5.0 100 $-3.5$ 99 [Nd/Fe] 3500 5500 $-1.0$ 5.5 2.0 100 $-3.5$ 99

Definition of the quality flag of the CN equivalent width difference with respect to the standard C and N abundances. CN_EW_err is the uncertainty of equivalent witdh of the cyanogen absorption line (cn0ew_gspspec_uncertainty); CN_p1 is the measured central wavelength of the cyanogen absorption line (cn0_gspspec_centralline); CN_p2 is the width of the cyoanogen line (cn0_gspspec_width):

 Flag name Condition Flag value DeltaCNq vbroadT$\geq$1 or vbroadG$\geq$1 or vbroadM$\geq$1 or CN_EW_err $=$ 0 or SNR$\leq$80 or gof$\geq-3.5$ or Filter CN $T_{\rm eff}$$\geq$4800 K or $\log g$$\geq$3.8 or Flag $=$ 9 abs(CN_p1 $-$ 862.884)$\geq$0.05 or CN_p2$\geq$0.25 vbroadT$<$1 and vbroadG$<$1 and vbroadM$<$1 and Flag $=$ 0 CN_EW$\_$err$\neq$0 and SNR$>$80 and gof$<-3.5$ and $T_{\rm eff}$$<$4800 K and $\log g$$<$3.8 and abs(CN_p1 $-$ 862.884)$<$0.05 and CN_p2$<$0.25

Definition of the quality flag for the diffuse interstellar band parameterisation. $p_{0}$ is the depth of the DIB (dibp0_gspspec); $p_{1}$ is the measured central wavelength of the DIB (dib_gspspec_lambda); $p_{2}$ is the width of the DIB (dibp2_gspspec); $R_{\rm a}$ is the standard deviation of the data–model residuals between 860.5 and 864.0 nm; $R_{\rm b}$ is the local noise level within the DIB profile :

 Flag name Condition Flag value DIBq SNR$\leq$50 or radial_velocity_error$>$5 km/s or $T_{\rm eff}$$<$3500 K or $T_{\rm eff}$$>$10${}^{6}$ K or Filter DIB $flux<$ 0 or Flag $=$ 9 $p_{0}$ $<$ 3/SNR $p_{1}<$ 861.66 nm or $p_{1}>$ 862.81 nm or Flag $=$ 5 $p_{0}>0.15$ 861.66 nm $ 862.81 nm and Flag $=$ 4 $p_{0} and $0.6 861.66 nm $ 862.81 nm and Flag $=$ 3 $p_{0}>R_{\rm b}$ and $0.6 861.66 nm $ 862.81 nm and Flag $=$ 2 $p_{0}>\max(R_{\rm a},R_{\rm b})$ and $0.6 861.66 nm $ 862.81 nm and Flag $=$ 1 $p_{0}>R_{\rm b}$ and $1.2 861.66 nm $ 862.81 nm and Flag $=$ 0 $p_{0}>\max(R_{\rm a},R_{\rm b})$ and $1.2

logchisq_gspspec : Logarithm of the goodness-of-fit for the GSP-Spec MatisseGauguin parameters (float)

Logarithm to the base 10 of the chi-squared between input spectrum rebinned to 800 pixels and solution synthetic spectrum computed from GSP-Spec MatisseGauguin (Recio-Blanco and et al. 2022).

ew_espels_halpha : Halpha pseudo-equivalent width from ESP-ELS (float, Length & Distance[nm])

Pseudo-equivalent width of the H$\alpha$ line measured on the RP spectra (Section 11.3.7). The value is expected to be negative when emission is present. To try to compensate for the existence of blends with species other than hydrogen in the cooler stars and assuming that no photospheric H$\alpha$ absorption is expected for K and M stars, we subtracted at $T_{\mathrm{eff}}<$ 5 000 K the pseudo-equivalent width measured on a synthetic spectrum with astrophysical parameters close to those derived by GSP-Phot for the target. The value that was subtracted (i.e. when $T_{\mathrm{eff}}<$ 5 000 K) is stored in ew_espels_halpha_model. ESP-ELS was only applied on targets brighter than magnitude G=17.65. More information on the module can be found in Section 11.3.7 and Section 11.4.4.

ew_espels_halpha_uncertainty : Uncertainty of the Halpha pseudo-equivalent width from ESP-ELS (float, Length & Distance[nm])

Uncertainty estimated on the pseudo-equivalent width of the H$\alpha$ line. It is computed by propagating the RP flux uncertainties through the integration over the considered H$\alpha$ window. Correlations between samples were ignored. ESP-ELS was only applied on targets brighter than magnitude G=17.65. More information on the module can be found in Section 11.3.7 and Section 11.4.4.

ew_espels_halpha_flag : Quality flag of the Halpha pseudo-equivalent width from ESP-ELS (string)

Quality flag of the H$\alpha$ pseudo-equivalent width. It takes the following values:

• 0: if the value stored in ew_espels_halpha_model was not subtracted ($T_{\mathrm{eff}}\geq$ 5 000 K).

• 1: if the value stored in ew_espels_halpha_model was subtracted ($T_{\mathrm{eff}}<$ 5 000 K).

ESP-ELS was only applied on targets brighter than magnitude G=17.65. More information on the module can be found in Section 11.3.7 and Section 11.4.4.

ew_espels_halpha_model : Halpha pseudo-equivalent width from ESP-ELS measured on the synthetic spectrum (float, Length & Distance[nm])

H$\alpha$ pseudo-equivalent width measured on the synthetic spectrum having the nearest astrophysical parameters to those derived by GSP-Phot (best library value, before post-processing). The value is available even when no ELS classification was provided and even when the value was not subtracted from the pseudo-equivalent width measured on the observed spectrum. ESP-ELS was only applied on targets brighter than magnitude G=17.65. More information on the module can be found in Section 11.3.7 and Section 11.4.4.

classlabel_espels : Adopted ELS class label from ESP-ELS (string)

The emission-line star class is based on the analysis of the BP/RP spectrum. Two random forest algorithms were used to classify the ELS targets. The first classifier is aimed to identify Wolf-Rayet stars and planetary nebulæ, and was trained on a subset of Gaia spectra chosen to be representative of each class. The second classifier is applied once significant H$\alpha$ emission was detected in order to identify Be, Herbig Ae/Be, T Tauri, and active M dwarf stars. It was trained on Gaia BP/RP spectra and on the astrophysical parameters. The astrophysical parameters adopted during the training and the processing are those obtained by GSP-Phot (best library value, taken before any post-processing filter or calibration was applied) for the target. The class label corresponds to the one having the highest probability and can take the following values:

• beStar (Be star)

• HerbigStar (Herbig Ae/Be star)

• PlanetaryNebula (Planetary Nebula)

• RedDwarfEmStar (active M dwarf star)

• TTauri (T Tauri star)

• wC (Wolf-Rayet star of type WC)

• wN (Wolf-Rayet star of type WN)

When no emission was found or when the spectrum could not be classified, the field is empty. ESP-ELS was only applied on targets brighter than magnitude G=17.65. More information on the module can be found in Section 11.3.7 and Section 11.4.4.

classlabel_espels_flag : Quality flag of the adopted ELS class label from ESP-ELS (string)

The quality flag provided with the ELS class label (classlabel_espels) indicates better quality for lower values of the flag. In a first instance, the quality assessment is based on the difference ($\Delta p$) between the 2 highest probabilities. The flag therefore takes the following values:

• $\Delta p\geq 0.8$: 0

• $\Delta p\geq 0.6$: 1

• $\Delta p\geq 0.4$: 2

• $\Delta p\geq 0.2$: 3

• $\Delta p<0.2$: 4

It is important to note that the identification of Be stars, Herbig Ae/Be stars, T Tauri stars, and active M dwarf stars also relies on the astrophysical parameters (APs) derived by GSP-Phot. After validation and during the post-processing, a significant fraction of APs that were suspected of being wrong or inaccurate have been removed or changed. On the other hand part of the APs that survived the post-processing disagree significantly with the spectal type tag provided by ESP-ELS and may point towards issues with the APs, the spectral type tag, or/and with the input data. To identify both cases, the quality flag was updated after processing as follows:

• classlabel_espels_flag+10: APs and spectral type tag are not consistent.

• classlabel_espels_flag+20: APs were removed during the post-processing.

ESP-ELS was only applied on targets brighter than magnitude G=17.65. More information on the module can be found in Section 11.3.7 and Section 11.4.4.

classprob_espels_wcstar : Probability from ESP-ELS of being a Wolf-Rayet star of type WC (float)

The probability of being a Wolf-Rayet star of type WC is derived on the BP/RP spectra by applying a random forest algorithm trained on a subset of Gaia spectra chosen to be representative of the WC stellar class. ESP-ELS was only applied on targets brighter than magnitude G=17.65. More information on the module can be found in Section 11.3.7 and Section 11.4.4.

classprob_espels_wnstar : Probability from ESP-ELS of being a Wolf-Rayet star of type WN (float)

The probability of being a Wolf-Rayet star of type WN is derived on the BP/RP spectra by applying a random forest algorithm trained on a subset of Gaia spectra chosen to be representative of the WN stellar class. ESP-ELS was only applied on targets brighter than magnitude G=17.65. More information on the module can be found in Section 11.3.7 and Section 11.4.4.

classprob_espels_bestar : Probability from ESP-ELS of being a Be Star (float)

The probability of being a Be star is derived on the BP/RP spectra by applying a random forest algorithm trained on a subset of Gaia spectra chosen to be representative of the Be stellar class, as well as on the corresponding astrophysical parameters provided by GSP-Phot (best library estimate before post-processing). ESP-ELS module was only applied on targets brighter than magnitude G=17.65. More information on the module can be found in Section 11.3.7 and Section 11.4.4.

classprob_espels_ttauristar : Probability from ESP-ELS of being a T Tauri Star (float)

The probability of being a T Tauri star is derived on the BP/RP spectra by applying a random forest algorithm trained on a subset of Gaia spectra chosen to be representative of the T Tauri stellar class, as well as on the corresponding astrophysical parameters provided by GSP-Phot (best library estimate before post-processing). ESP-ELS was only applied on targets brighter than magnitude G=17.65. More information on the module can be found in Section 11.3.7 and Section 11.4.4.

classprob_espels_herbigstar : Probability from ESP-ELS of being a Herbig Ae/Be Star (float)

The probability of being a Herbig Ae/Be star is derived on the BP/RP spectra by applying a random forest algorithm trained on a subset of Gaia spectra chosen to be representative of the Herbig Ae/Be stellar class, as well as on the corresponding astrophysical parameters provided by GSP-Phot (best library estimate before post-processing). ESP-ELS was only applied on targets brighter than magnitude G=17.65. More information on the module can be found in Section 11.3.7 and Section 11.4.4.

classprob_espels_dmestar : Probability from ESP-ELS of being an active M dwarf Star (float)

The probability of being an active M dwarf (dMe) star is derived on the BP/RP spectra by applying a random forest algorithm trained on a subset of Gaia spectra chosen to be representative of the dMe stellar class, as well as on the corresponding astrophysical parameters provided by GSP-Phot (best library estimate before post-processing). ESP-ELS was only applied on targets brighter than magnitude G=17.65. More information on the module can be found in Section 11.3.7 and Section 11.4.4.

classprob_espels_pne : Probability from ESP-ELS of being a planetary nebula (float)

The probability of being a planetary nebula (PN/PNe) is derived on the BP/RP spectra by applying a random forest algorithm trained on a subset of Gaia spectra chosen to be representative of PNe. ESP-ELS was only applied on targets brighter than magnitude G=17.65. More information on the module can be found in Section 11.3.7 and Section 11.4.4.

azero_esphs : Monochromatic interstellar extinction, A${}_{\mathrm{0}}$, from ESP-HS (float, Magnitude[mag])

Monochromatic interstellar extinction, A${}_{\mathrm{0}}$, derived at 541.4 nm by ESP-HS from the comparison of the observed BP/RP and, when available, RVS spectra to simulations. ESP-HS is processing targets brighter than magnitude G=17.65, and which are tagged O, B, or A in the spectraltype_esphs field. A more detailed description of the module is provided in Section 11.3.8 and Section 11.4.6.

azero_esphs_uncertainty : Uncertainty at a 68% confidence level on A${}_{\mathrm{0}}$ from ESP-HS (float, Magnitude[mag])

Uncertainty on A${}_{\mathrm{0}}$ derived by ESP-HS. The value is provided by the diagonal of the covariance matrix. At all steps, the correlation between flux samples is ignored. Uncertainties were found to be underestimated by a factor 5 to 10 in the BP/RP+RVS processing mode (i.e. first digit/character of flags_esphs has value 0). ESP-HS is processing targets brighter than magnitude G=17.65, and which are tagged O, B, or A in the spectraltype_esphs field. A more detailed description of the module is provided in Section 11.3.8 and Section 11.4.6.

ag_esphs : Intersterstellar extinction in G band from ESP-HS (float, Magnitude[mag])

Interstellar extinction in G band, $A_{\rm G}$, derived by ESP-HS from the comparison of the observed BP/RP and, when available, RVS spectra to simulations. ESP-HS is processing targets brighter than magnitude G=17.65, and which are tagged O, B, or A in the spectraltype_esphs field. A more detailed description of the module is provided in Section 11.3.8 and Section 11.4.6.

ag_esphs_uncertainty : Uncertainty on $A_{\rm G}$ from ESP-HS (float, Magnitude[mag])

Uncertainty on $A_{\rm G}$ derived by ESP-HS. The value is inferred from the uncertainty of A${}_{\mathrm{0}}$. Uncertainties were found to be underestimated by a factor 5 to 10 in the BP/RP+RVS processing mode (i.e. first digit/character of flags_esphs has value 0). ESP-HS is processing targets brighter than magnitude G=17.65, and which are tagged O, B, or A in the spectraltype_esphs field. A more detailed description of the module is provided in Section 11.3.8 and Section 11.4.6.

ebpminrp_esphs : Reddening $E(G_{\rm BP}-G_{\rm RP})$ from ESP-HS (float, Magnitude[mag])

Interstellar reddening, $E(G_{\rm BP}-G_{\rm RP})$, derived by ESP-HS from the comparison of the observed BP/RP and, when available, RVS spectra to simulations. ESP-HS is processing targets brighter than magnitude G=17.65, and which are tagged O, B, or A in the spectraltype_esphs field. A more detailed description of the module is provided in Section 11.3.8 and Section 11.4.6.

ebpminrp_esphs_uncertainty : Uncertainty on $E(G_{\rm BP}-G_{\rm RP})$ from ESP-HS (float, Magnitude[mag])

Uncertainty on $E(G_{\rm BP}-G_{\rm RP})$ derived by ESP-HS. The value is inferred from the uncertainty of A${}_{\mathrm{0}}$. Uncertainties were found to be underestimated by a factor 5 to 10 in the BP/RP+RVS processing mode (i.e. first digit/character of flags_esphs has value 0). ESP-HS is processing targets brighter than magnitude G=17.65, and which are tagged O, B, or A in the spectraltype_esphs field. A more detailed description of the module is provided in Section 11.3.8 and Section 11.4.6.

teff_esphs : Effective temperature from ESP-HS (float, Temperature[K])

Effective temperature derived by fitting the BP/RP and, when available, the RVS spectra with synthetic spectra. The module assumes a solar chemical composition. ESP-HS is processing targets brighter than magnitude G=17.65, and which are tagged O, B, or A in the spectraltype_esphs field. A more detailed description of the module is provided in Section 11.3.8 and Section 11.4.4.

teff_esphs_uncertainty : Uncertainty at a 68% confidence level on the effective temperature from ESP-HS (float, Temperature[K])

Uncertainty on the effective temperature derived by ESP-HS. The value is extracted from the diagonal of the covariance matrix. At all steps, the correlation between flux samples is ignored. Uncertainties were found to be underestimated by a factor 5 to 10 in the BP/RP+RVS processing mode (i.e. first digit/character of spectraltype_esphs has value 0). A more detailed description of ESP-HS is provided in Section 11.3.8 and Section 11.4.4.

logg_esphs : Surface gravity from ESP-HS (float, GravitySurface[log cgs])

Surface gravity derived by fitting the BP/RP and, when available, the RVS spectra with synthetic spectra. The module assumes a solar chemical composition. ESP-HS is processing targets brighter than magnitude G=17.65, and which are tagged O, B, or A in the spectraltype_esphs field. A more detailed description of the module is provided in Section 11.3.8 and Section 11.4.4.

logg_esphs_uncertainty : Uncertainty at a 68% confidence level on the surface gravity from ESP-HS (float, GravitySurface[log cgs])

Uncertainty on the surface gravity derived by ESP-HS. The value is extracted from the diagonal of the covariance matrix. At all steps, the correlation between flux samples is ignored. Uncertainties were found to be underestimated by a factor 5 to 10 in the BP/RP+RVS processing mode (i.e. first digit/character of flags_esphs has value 0). A more detailed description of ESP-HS is provided in Section 11.3.8 and Section 11.4.4.

vsini_esphs : Projected rotational velocity from ESP-HS (float, Velocity[km s${}^{-1}$])

The line broadening of the RVS spectrum is derived by assuming that it is due to stellar rotation and by adopting the same method as described in Frémat et al. (2022). Therefore, we named it $v\sin i$. A value is only provided in the BP/RP+RVS mode (i.e. first digit/character of flags_esphs has value 0). ESP-HS is processing targets brighter than magnitude G=17.65, and which are tagged O, B, or A in the spectraltype_esphs field. A more detailed description of the module is provided in Section 11.3.8 and Section 11.4.4.

vsini_esphs_uncertainty : Uncertainty on the projected rotational velocity from ESP-HS (float, Velocity[km s${}^{-1}$])

The uncertainty on $v\sin i$ was derived by adopting the approach described in Zucker (2003).

flags_esphs : Quality flag of the ESP-HS parametrisation (string)

The quality flag usually has 2 digits. The first digit tells what processing mode was used to derive the effective temperature, surface gravity, interstellar extinction and reddening, and $v\sin i$. It takes the following values:

• 0: when both BP/RP and RVS spectra were used.

• 1: when BP/RP only was used. In this mode no $v\sin i$ is available.

The second digit is relative to the spectral type (spectraltype_esphs) that is derived from the analysis of the BP/RP spectrum only. During the processing a probability was assigned, and used for the quality assessment of the spectral type tagging. Therefore, the second digit of flags_esphs ranges from 1 to 5, depending on the first (p1) and second (p2) highest probability as follows:

• p1 $>0.5$ and p2 $\leq 0.1$: second digit of flags_esphs = 1

• p1 $>0.5$ and p2 $\leq 0.2$: second digit of flags_esphs = 2

• p1 $>0.5$ and p2 $\leq 0.3$: second digit of flags_esphs = 3

• p1 $>0.5$ and p2 $\leq 0.4$: second digit of flags_esphs = 4

• p1 $\leq 0.5$: second digit of flags_esphs = 5

• target brighter than G=17.65, but no spectral type tag was derived: flags_esphs = 999

During the validation of the CSTAR spectral type tag used to identify candidate carbon stars (Gaia Collaboration et al. 2022c), it was noted that only a fraction of these had significantly stronger than normal C${}_{2}$ and CN molecular bands. We flagged these targets by setting the second digit of flags_esphs to 0.

When the algorithm was not able to properly set the quality flag (for 2.7 % of the targets with a spectral type tag) it received the value ’999’. In these cases, no parameters nor classification was derived but the corresponding spectral type tag was kept.

spectraltype_esphs : Spectral type from ESP-ELS (string)

The spectral type tag is obtained by ESP-ELS. At the origin it was obtained by ESP-HS (we kept the module name), but the corresponding algorithm was later moved to the upstream module ESP-ELS. It is derived from the analysis of the BP/RP spectrum only, and takes the following values: CSTAR,M,K,G,F,A,B,O. ESP-ELS is processing targets brighter than magnitude G=17.65. A more detailed description of the module is provided in Section 11.3.7 and Section 11.4.4.

activityindex_espcs : Chromospheric activity index from ESP-CS, measured on the calcium triplet using RVS spectra (float, Length & Distance[nm])

The activity index from the Apsis module ESP-CS is computed by comparing the observed RVS spectrum with a purely photospheric template spectrum obtained by interpolating in a grid of synthetic spectra. The atmospheric parameters (APs) adopted in the interpolation are taken from the output of either GSP-Spec or GSP-Phot. GSP-Spec APs are adopted when all three parameters teff_gspspec, logg_gspspec, and mh_gspspec are provided. Otherwise teff_gspphot, logGspphot, and mh_gspphot are adopted, if they are all provided. The activity index gives the excess equivalent width factor in the cores of the Ca II infrared triplet lines with respect to the template inactive spectrum. The excess equivalent width is computed for each of the Ca II infrared triplet lines and the three values obtained are averaged. When the projected rotational velocity vbroad in table gaia_source is provided, rotational broadening is taken into account. See Section 11.3.9 for details.

activityindex_espcs_uncertainty : Uncertainty in the chromospheric activity index from ESP-CS (float, Length & Distance[nm])

Uncertainty in the activity index from ESP-CS (activityindex_espcs). The uncertainty is computed by considering the standard deviation of the excess equivalent width factor in each of the Ca II infrared triplet lines, taking spectrum noise into account, and applying error propagation. See Section 11.3.9 for details.

activityindex_espcs_input : Source of input stellar parameters for the computation of the activity index by ESP-CS (string)

Flag indicating the source of the stellar atmospheric parameters used by ESP-CS in deriving the activity index. The flag has the value “M1” if the source is GSP-Spec, and “M2” if the source is GSP-Phot. See the description of activityindex_espcs.

teff_espucd : Effective temperature estimate from ESP-UCD based on the RP spectrum (float, Temperature[K])

Effective temperature estimate from ESP-UCD inferred from the RP spectrum. The prediction module is based on a Gaussian Process trained with empirical examples. See Section 11.3.10 for details.

teff_espucd_uncertainty : Uncertainty of the effective temperature estimate produced by ESP-UCD (float, Temperature[K])

Standard deviation of 10 effective temperature predictions obtained by the ESP-UCD module for 10 RP spectra generated using random sampling from Gaussian distributions centred at the observed fluxes and with standard deviations given by the flux uncertainties. The value thus obtained is rescaled (multiplied by 7) to match the root-mean-square error of the module predictions for a set of well-known ultracool dwarfs (see Section 11.3.10 for further details).

flags_espucd : Quality flags of the ESP-UCD parameter estimates (string)

Two-digits ESP-UCD parameters quality flag. The first digit (0, 1 or 2) is based on the goodness-of-fit estimate and the RP spectrum signal-to-noise ratio, 0 being the best quality. The second digit is 1 for sources with inconsistent $T_{\rm eff}$ predictions given the value of $G+5\cdot\log_{10}(\varpi)+5$. It is 0 for all other sources. See Section 11.3.10 for a quantitative definition of the three quality categories for the first digit and of the inconsistency criterion for the second digit.

radius_flame : Radius of the star from FLAME using teff_gspphot and lum_flame (float, Length & Distance[Solar Radius])

The radius of the star from FLAME, derived from teff_gspphot and lum_flame using the Stefan-Boltzmann law with a solar effective temperature of 5772 K, see Section 1.2.3, and associated uncertainties. It is defined as the median value (50${}^{th}$ percentile) of the distribution from sampling.

Lower confidence level (16%) of the radius of the star from FLAME, see description for radius_flame. It is derived from teff_gspphot, lum_flame and associated uncertainties. It is defined as the 16${}^{th}$ percentile value of the distribution from sampling. Upper and lower levels include 68% confidence interval.

Upper confidence level (84%) of the radius of the star from FLAME, see description for radius_flame. It is derived from teff_gspphot, lum_flame and associated uncertainties. It is defined as the 84${}^{th}$ percentile value of the distribution from sampling. Upper and lower levels include 68% confidence interval.

lum_flame : Luminosity of the star from FLAME using G band magnitude, extinction (ag_gspphot), parallax or distance, and a bolometric correction bc_flame (float, Luminosity[Solar Luminosity])

Luminosity of the star from FLAME using G band magnitude, extinction from GSP-Phot (ag_gspphot), parallax or distance_gspphot, and a bolometric correction bc_flame. It is defined as the median value (50${}^{th}$ percentile) of the distribution from sampling. The bolometric correction depends on the effective temperature, metallicity and surface gravity, and these are based on GSP-Phot values. The bolometric magnitude of the Sun = 4.74 mag, see Section 1.2.3 and the reference absolute G-band magnitude of the Sun is 4.66 mag, see Section 11.3.6.

lum_flame_lower : Lower confidence level (16%) of lum_flame (float, Luminosity[Solar Luminosity])

Lower confidence level (16%) of the luminosity of the star from FLAME, see description for lum_flame. It is defined as the 16${}^{th}$ percentile value of the distribution from sampling. Upper and lower levels contain the 68% confidence interval.

lum_flame_upper : Upper confidence level (84%) of lum_flame (float, Luminosity[Solar Luminosity])

The upper confidence level (84%) of the luminosity of the star from FLAME, see description for lum_flame. It is defined as the 84${}^{th}$ percentile value of the distribution from sampling. Upper and lower levels contain the 68% confidence interval.

mass_flame : Mass of the star from FLAME using stellar models, lum_flame, and teff_gspphot (float, Mass[Solar Mass])

Mass of the star from FLAME. It is defined as the median value (50${}^{th}$ percentile) of the 1D projected distribution from sampling in mass and age. It is derived by comparing teff_gspphot and lum_flame to the BASTI solar metallicity stellar evolution models (Hidalgo et al. 2018), see Section 11.3.6 for details.

mass_flame_lower : Lower confidence level (16%) of mass_flame (float, Mass[Solar Mass])

Lower confidence level (16%) of the mass of the star from FLAME, see description for mass_flame. It is defined as the 16${}^{th}$ percentile value of the 1D projected distribution from sampling in mass and age. Upper and lower levels contain the 68% confidence interval.

mass_flame_upper : Upper confidence level (84%) of mass_flame (float, Mass[Solar Mass])

Upper confidence level (84%) of the mass of the star from FLAME, see description for mass_flame. It is defined as the 84${}^{th}$ percentile value of the 1D projected distribution from sampling in mass and age. Upper and lower levels contain the 68% confidence interval.

age_flame : Age of the star from FLAME using stellar models, see mass_flame for details (float, Time[Gyr])

Age of the star from FLAME. It is defined as the median value (50${}^{th}$ percentile) of the 1D projected distribution from sampling in mass and age, see mass_flame for details.

age_flame_lower : Lower confidence level (16%) of age_flame (float, Time[Gyr])

Lower confidence level (16%) of the age of the star from FLAME, see description for age_flame. It is defined as the 16${}^{th}$ percentile value of the 1D projected distribution from sampling in mass and age. Upper and lower levels contain the 68% confidence interval.

age_flame_upper : Upper confidence level (84%) of age_flame (float, Time[Gyr])

Upper confidence level (84%) of the age of the star from FLAME, see description for age_flame. It is defined as the 84${}^{th}$ percentile value of the 1D projected distribution from sampling in mass and age. Upper and lower levels contain the 68% confidence interval.

flags_flame : Flags indicating quality and processing information from FLAME (string)

This field contains the quality and processing flags from FLAME and takes the form ‘AB’. The first digit refers to the quality of the mass and age determination (A = 0, 1, 2) and the second digit (B) informs the user if the parallax or distance_gspphot was used for deriving all FLAME stellar parameters (lum_flame, radius_flame,…). Concerning the quality of the mass and age (first digit, A), the following is adopted:

• A=0

nothing to report

• A=1

the star is a giant and the mass and age should be considered with an uncertainty on the order of 20% - 30%

• A=2

the mass and age are not available

We note that while the evolstage_flame parameter is related to mass and age, this flag is not applicable to this parameter.

For the second digit, B, the following is adopted:

• B=0

parallax is used

• B=1

distance_gspphot is used

• B=2

parallax is used due to convergence issues with distance_gspphot

See Section 11.3.6 of the online documentation for details.

evolstage_flame : Evolutionary stage of the star from FLAME using stellar models, see mass_flame for details (int)

Evolution stage of the star from FLAME. It is an integer value typically between 100 and 1300 and defined using the median value (50${}^{th}$ percentile) of the 1D projected distribution from sampling in mass and age, see mass_flame for details. The value is adapted from the BASTI model grid (Hidalgo et al. 2018) adopting the following convention:

• $100=$ zero age main sequence (ZAMS)

• $300=$ first minimum of $T_{\rm eff}$ for massive stars or central hydrogen mass fraction = 0.30 for low-mass stars

• $360=$ main sequence turn-off

• $420=$ central hydrogen mass fraction = 0.00

• $490=$ base of the red giant branch (RGB)

• $860=$ maximum $\cal{L}$ along the RGB bump

• 890 = minimum $\cal{L}$ along the RGB bump

• 1290 = tip of the RGB

gravredshift_flame : Gravitational redshift from FLAME using radius_flame and logg_gspphot (float, Velocity[km s${}^{-1}$])

Gravitational redshift, in velocity, from FLAME using radius_flame and logg_gspphot.

gravredshift_flame_lower : Lower confidence level (16%) of gravredshift_flame (float, Velocity[km s${}^{-1}$])

Lower confidence level (16%) of the gravitational redshift of the star from FLAME, see description for gravredshift_flame. It is defined as the 16${}^{th}$ percentile value of the distribution from sampling. Upper and lower levels contain the 68% confidence interval.

gravredshift_flame_upper : Upper confidence level (84%) of gravredshift_flame (float, Velocity[km s${}^{-1}$])

Upper confidence level (84%) of the gravitational redshift of the star from FLAME, see description for gravredshift_flame. It is defined as the 84${}^{th}$ percentile value of the distribution from sampling. Upper and lower levels contain the 68% confidence interval.

bc_flame : Bolometric correction used to derive lum_flame (float, Magnitude[mag])

Bolometric correction for the G-band magnitude (${\mathrm{BC}_{G}}$) used to derive lum_flame. It is defined as the median value (50${}^{th}$ percentile) of the distribution from sampling. It is a function of effective temperature, surface gravity, and metallicity, and has been derived from MARCS models, see Section 11.2.3 of the online documentation. The bolometric correction for the Sun is defined as +0.08 mag, see Section 11.3.6, where M${}_{\rm bol,\odot}=4.74$ mag, see Section 1.2.3, i.e. M${}_{G\odot}=4.66$ mag.

mh_msc : Metallicity of the source treated as a binary system from MSC using BP/RP spectra and parallax (float, Abundances[dex])

Decimal logarithm of the ratio of the average number abundance of elements heavier than helium compared to hydrogen relative to the same ratio of solar abundances ([M/H]) (assuming source is a binary system) from MSC using BP/RP spectra and parallax. Because MSC uses an empirical BPRP forward model trained on APOGEE astrophysical parameters the results are tied to the metallicity scale of the set of APOGEE targets used as the training set. The metallicity value is the median of the MCMC samples. It is assumed that both components of the binary system have the same metallicity. For details see Section 11.3.5.

mh_msc_upper : Upper confidence level (84%) of the metallicity from MSC using BP/RP spectra and parallax (float, Abundances[dex])

Upper confidence level of the metallicity inferred by MSC from BP/RP spectra and parallax. This is the 84th percentile. Lower and upper levels include 68% confidence (corresponding to a conventional 1-sigma interval).

mh_msc_lower : Lower confidence level (16%) of the metallicity from MSC using BP/RP spectra and parallax (float, Abundances[dex])

Lower confidence level of the metallicity inferred by MSC from BP/RP spectra and parallax. This is the 16th percentile. Lower and upper levels include 68% confidence (corresponding to a conventional 1-sigma interval).

azero_msc : Monochromatic extinction $A_{0}$ at 541.4 nm of the source treated as a binary system from MSC using BP/RP spectra and parallax (float, Magnitude[mag])

Monochromatic extinction $A_{0}$ at 541.4 nm of the source (assuming source is a binary system) inferred by MSC from BP/RP spectra and parallax. This is the median of the MCMC samples. NB: This is the extinction parameter in the adopted Fitzpatrick extinction law (Fitzpatrick 1999, see Section 11.2.3 of the online documentation).

azero_msc_upper : Upper confidence level (84%) of monochromatic extinction $A_{0}$ at 541.4 nm from MSC using BP/RP spectra and parallax (float, Magnitude[mag])

Upper confidence level of $A_{0}$ inferred by MSC from BP/RP spectra and parallax. This is the 84th percentile. Lower and upper levels include 68% confidence (corresponding to a conventional 1-sigma interval).

azero_msc_lower : Lower confidence level (16%) of monochromatic extinction $A_{0}$ at 541.4 nm from MSC using BP/RP spectra and parallax (float, Magnitude[mag])

Lower confidence level of $A_{0}$ inferred by MSC from BP/RP spectra and parallax. This is the 16th percentile. Lower and upper levels include 68% confidence (corresponding to a conventional 1-sigma interval).

distance_msc : Distance from MSC using BP/RP spectra and parallax (float, Length & Distance[pc])

Distance of the source (assuming source is a binary system) inferred by MSC from BP/RP spectra and parallax. This is the median of the MCMC samples. For details see Section 11.3.5.

distance_msc_upper : Upper confidence level (84%) of distance from MSC using BP/RP spectra and parallax (float, Length & Distance[pc])

Upper confidence level of distance inferred by MSC from BP/RP spectra and parallax. This is the 84th percentile. Lower and upper levels include 68% confidence (corresponding to a conventional 1-sigma interval).

distance_msc_lower : Lower confidence level (16%) of distance from MSC using BP/RP spectra and parallax (float, Length & Distance[pc])

Lower confidence level of distance inferred by MSC from BP/RP spectra and parallax. This is the 16th percentile. Lower and upper levels include 68% confidence (corresponding to a conventional 1-sigma interval).

teff_msc1 : Effective temperature of the primary from MSC using BP/RP spectra and parallax (float, Temperature[K])

Effective temperature of the primary (assuming source is a binary system and the primary is the component with more flux in the BP and RP spectra combined) inferred by MSC from BP/RP spectra and parallax. This is the median of the MCMC samples. For details see Section 11.3.5.

teff_msc1_upper : Upper confidence level (84%) of effective temperature of the primary from MSC using BP/RP spectra and parallax (float, Temperature[K])

Upper confidence level of effective temperature of the primary inferred by MSC from BP/RP spectra and parallax. This is the 84th percentile. Lower and upper levels include 68% confidence (corresponding to a conventional 1-sigma interval).

teff_msc1_lower : Lower confidence level (16%) of effective temperature of the primary from MSC using BP/RP spectra and parallax (float, Temperature[K])

Lower confidence level of effective temperature of the primary inferred by MSC from BP/RP spectra and parallax. This is the 16th percentile. Lower and upper levels include 68% confidence (corresponding to a conventional 1-sigma interval).

teff_msc2 : Effective temperature of the secondary from MSC using BP/RP spectra and parallax (float, Temperature[K])

Effective temperature of the secondary (assuming source is a binary system and the secondary is the component with less flux in the BP and RP spectra combined) inferred by MSC from BP/RP spectra and parallax. This is the median of the MCMC samples. For details see Section 11.3.5.

teff_msc2_upper : Upper confidence level (84%) of effective temperature of the secondary from MSC using BP/RP spectra and parallax (float, Temperature[K])

Upper confidence level of effective temperature of the secondary inferred by MSC from BP/RP spectra and parallax. This is the 84th percentile. Lower and upper levels include 68% confidence (corresponding to a conventional 1-sigma interval).

teff_msc2_lower : Lower confidence level (16%) of effective temperature of the secondary from MSC using BP/RP spectra and parallax (float, Temperature[K])

Lower confidence level of effective temperature of the secondary inferred by MSC from BP/RP spectra and parallax. This is the 16th percentile. Lower and upper levels include 68% confidence (corresponding to a conventional 1-sigma interval).

logg_msc1 : Surface gravity of the primary from MSC using BP/RP spectra and parallax (float, GravitySurface[log cgs])

Surface gravity of the primary (assuming source is a binary system and the primary is the component with more flux in the BP and RP spectra combined) inferred by MSC from BP/RP spectra and parallax. This is the median of the MCMC samples. For details see Section 11.3.5.

logg_msc1_upper : Upper confidence level (84%) of surface gravity of the primary from MSC using BP/RP spectra and parallax (float, GravitySurface[log cgs])

Upper confidence level of surface gravity of the primary inferred by MSC from BP/RP spectra and parallax. This is the 84th percentile. Lower and upper levels include 68% confidence (corresponding to a conventional 1-sigma interval).

logg_msc1_lower : Lower confidence level (16%) of surface gravity of the primary from MSC using BP/RP spectra and parallax (float, GravitySurface[log cgs])

Lower confidence level of surface gravity of the primary inferred by MSC from BP/RP spectra and parallax. This is the 16th percentile. Lower and upper levels include 68% confidence (corresponding to a conventional 1-sigma interval).

logg_msc2 : Surface gravity of the secondary from MSC using BP/RP spectra and parallax (float, GravitySurface[log cgs])

Surface gravity of the secondary (assuming source is a binary system and the secondary is the component with less flux in the BP and RP spectra combined) inferred by MSC from BP/RP spectra and parallax. This is the median of the MCMC samples. For details see Section 11.3.5.

logg_msc2_upper : Upper confidence level (84%) of surface gravity of the secondary from MSC using BP/RP spectra and parallax (float, GravitySurface[log cgs])

Upper confidence level of surface gravity of the secondary inferred by MSC from BP/RP spectra and parallax. This is the 84th percentile. Lower and upper levels include 68% confidence (corresponding to a conventional 1-sigma interval).

logg_msc2_lower : Lower confidence level (16%) of surface gravity of the secondary from MSC using BP/RP spectra and parallax (float, GravitySurface[log cgs])

Lower confidence level of surface gravity of the secondary inferred by MSC from BP/RP spectra and parallax. This is the 16th percentile. Lower and upper levels include 68% confidence (corresponding to a conventional 1-sigma interval).

ag_msc : Extinction in G band of the source treated as a binary system from MSC using BP/RP spectra and parallax (float, Magnitude[mag])

G band extinction of the source (assuming source is a binary system) inferred by MSC from BP/RP spectra and parallax. This is the median of the MCMC samples. For details see Section 11.3.5.

ag_msc_upper : Upper confidence level (84%) of extinction in G band from MSC using BP/RP spectra and parallax (float, Magnitude[mag])

Upper confidence level of G band extinction inferred by MSC from BP/RP spectra and parallax. This is the 84th percentile. Lower and upper levels include 68% confidence (corresponding to a conventional 1-sigma interval).

ag_msc_lower : Lower confidence level (16%) of extinction in G band from MSC using BP/RP spectra and parallax (float, Magnitude[mag])

Lower confidence level of G band extinction inferred by MSC from BP/RP spectra and parallax. This is the 16th percentile. Lower and upper levels include 68% confidence (corresponding to a conventional 1-sigma interval).

logposterior_msc : Goodness-of-fit score (normalised log-posterior) of MSC MCMC (float)

Goodness-of-fit score of MSC MCMC. This was calculated as the mean log-posterior of all MCMC samples normalised with the uncertainty of the data. A higher value corresponds to a better fit.

mcmcaccept_msc : Mean MCMC acceptance rate of MSC MCMC (float)

Mean acceptance rate of MSC MCMC chain for all walkers.

mcmcdrift_msc : Mean drift of the MSC MCMC chain in units of parameter standard deviation (float)

Drift of the MCMC chain in units of parameter standard deviation, averaged over all parameters. Computed as the mean value of each parameter in the first MCMC ensemble state minus the mean value in the last MCMC ensemble state divided by the standard deviation of values in the last MCMC ensemble state. This is then averaged over all MSC parameters to provide a single value. If the MCMC chain has not converged, the first and last ensemble states will have low or zero overlap. In such a case, their mean values will show a large difference, larger than the standard deviation in the final ensemble state. Therefore, large values of this quantity indicate poor MCMC convergence, whereas values close to zero indicate good MCMC convergence.

flags_msc : Flag indicating quality information from MSC (string)

Catalogue flag for MSC. This is set to ‘0’ if logposterior_msc $\geq-1000$ & mcmcdrift_msc $\leq$ 1. In all other cases it is set to ‘1’ indicating an unreliable inference result.

neuron_oa_id : Identifier of the OA SOM map neuron that represents the source (long)

A unique identifier for the neuron that represents the source in the SOM map produced by the OA module. If the source was not considered as an outlier according to DSC classification or if it was discarded by OA module, then this field will be null. See Section 11.3.12 for further details.

neuron_oa_dist : Distance between the source XP spectra and the OA neuron XP prototype that represents the source (float)

Squared Euclidean distance between the source XP spectra (preprocessed) and the neuron XP prototype (xp_spectrum_prototype_flux in table oa_neuron_xp_spectra) that represents such a source. If the source was not considered as an outlier according to DSC classification or if it was discarded by the OA module, then this field will be null. See Section 11.3.12 for further details.

neuron_oa_dist_percentile_rank : Percentile rank according to the distance distribution of the OA neuron that represents the source (int)

Percentile rank according to the squared Euclidean distance distribution of the OA neuron that represents the source. If the source was not considered an outlier according to DSC classification or if it was discarded by the OA module, then this field will be null. See Section 11.3.12 for further details.

flags_oa : Flags indicating quality and processing information from OA (string)

Processing flags related to the quality of the classification and source processing performed by the OA module, which is encoded as follows:

 $ABCD$

where:

 Code Description Value range $A$ Number of gaps with negative fluxes in BP spectrum $[0,2]$ * $B$ Number of gaps with negative fluxes in RP spectrum $[0,2]$ * $C$ Source classification quality category $[0,5]$ $D$ Neuron quality category $[0,6]$

(*) $A$ and $B$ flags taking value 2 means two or more gaps taking negative flux values.