J/A+A/690/A54 SpectroTranslator GALAH and APOGEE examples (Thomas+, 2024)
SpectroTranslator:
Deep-neural network algorithm for homogenising spectroscopic parameters.
Thomas G.F., Battaglia G., Gran F., Fernandez-Alvar E., Tsantaki M.,
Pancino E., Hill V., Kordopatis G., Gallart C., Turchi A., Masseron T.
<Astron. Astrophys. 690, A54 (2024)>
=2024A&A...690A..54T 2024A&A...690A..54T (SIMBAD/NED BibCode)
ADC_Keywords: Models ; Abundances ; Effective temperatures
Keywords: methods: data analysis - techniques: spectroscopic - catalogs -
stars: abundances - stars: fundamental parameters - Galaxy: abundances
Abstract:
In modern Galactic astronomy, stellar spectroscopy plays a ivotal role
in complementing large photometric and astrometric surveys and
enabling deeper insights to be gained into the chemical evolution and
chemo-dynamical mechanisms at play in the Milky Way and its
satellites. Nonetheless, the use of different instruments and
dedicated pipelines in various spectroscopic surveys can lead to
differences in the derived spectroscopic parameters.
Efforts to homogenise these surveys onto a common scale are essential
to maximising their scientific legacy. To this aim, we developed the
SpectroTranslator, a data-driven deep neural network algorithm that
converts spectroscopic parameters from the base of one survey (base A)
to that of another (base B).
SpectroTranslator is comprised of two neural networks: an intrinsic
network, where all the parameters play a role in computing the
transformation, and an extrinsic network, where the outcome for one of
the parameters depends on all the others, but not the reverse. The
algorithm also includes a method to estimate the importance that the
various parameters play in the conversion from base A to B.
To demonstrate the workings of the algorithm, we applied it to
transform effective temperature, surface gravity, metallicity,
[Mg/Fe], and line-of-sight velocity from the base of GALAH DR3 into
the APOGEE- 2 DR17 base. We demonstrate the efficiency of the
SpectroTranslator algorithm to translate the spectroscopic parameters
from one base to another, directly using parameters by the survey
teams. We were able to achieve a similar performance than previous
works that have performed a similar type of conversion but using the
full spectrum, rather than the spectroscopic parameters. This allowed
us to reduce the computational time and use the output of pipelines
optimised for each survey. By combining the transformed GALAH
catalogue with the APOGEE- 2 catalogue, we studied the distribution of
[Fe/H] and [Mg/Fe] across the Galaxy and we found that the median
distribution of both quantities present a vertical asymmetry at large
radii. We attribute it to the recent perturbations generated by the
passage of a dwarf galaxy across the disc or by the infall of the
Large Magellanic Cloud.
Several aspects still need to be refined, such as the question of the
optimal way to deal with regions of the parameter space meagrely
populated by stars in the training sample. However, SpectroTranslator
has already demonstrated its capability and is poised to play a crucial
role in standardising various spectroscopic surveys onto a unified
framework.
Description:
Catalogues of translated parameters by the SpectroTranslator.
File Summary:
--------------------------------------------------------------------------------
FileName Lrecl Records Explanations
--------------------------------------------------------------------------------
ReadMe 80 . This file
g2a.dat 924 588464 GALAH DR3 catalogue transformed onto the
APOGEE-2 DR17 base
a2g.dat 913 675857 APOGEE-2 DR17 catalogue transformed onto the
GALAH DR3 base
--------------------------------------------------------------------------------
See also:
I/355 : Gaia DR3 Part 1. Main source (Gaia Collaboration, 2022)
III/286 : APOGEE-2 DR17 final allStar catalog (Abdurro'uf+, 2022)
J/MNRAS/506/150 : The GALAH+ Survey DR3 (Buder+, 2021)
Byte-by-byte Description of file: g2a.dat
--------------------------------------------------------------------------------
Bytes Format Units Label Explanations
--------------------------------------------------------------------------------
1- 15 I15 --- GALAH GALAH observation identifier (sobject_id)
17- 35 I19 --- GaiaDR3 Gaia DR3 unique source identifier
(source_id)
37- 55 E19.14 K Teff5 5th percentile of the SpectroTranslator
effective temperature (teffpred5)
57- 77 E21.17 K Teff16 16th percentile of the SpectroTranslator
effective temperature (teffpred16)
79- 99 E21.17 K Teff50 50th percentile of the SpectroTranslator
effective temperature (teffpred50)
101-124 F24.15 K Teff84 84th percentile of the SpectroTranslator
effective temperature (teffpred84)
126-150 F25.16 K Teff95 95th percentile of the SpectroTranslator
effective temperature (teffpred95)
152-177 F26.18 K e_Teff50 Systematic error of the SpectroTranslator on
the transformation of the effective
temperature (teffpredERR)
179-199 E21.18 [cm/s2] logg5 5th percentile of the SpectroTranslator
surface gravity (loggpred5)
201-222 E22.18 [cm/s2] logg16 16th percentile of the SpectroTranslator
surface gravity (loggpred16)
224-247 F24.18 [cm/s2] logg50 50th percentile of the SpectroTranslator
surface gravity (loggpred50)
249-271 F23.17 [cm/s2] logg84 84th percentile of the SpectroTranslator
surface gravity (loggpred84)
273-295 F23.17 [cm/s2] logg95 95th percentile of the SpectroTranslator
surface gravity (loggpred95)
297-317 E21.19 [cm/s2] e_logg50 Systematic error of the SpectroTranslator on
the surface gravity (loggpredERR)
319-340 E22.19 --- [Fe/H]5 5th percentile of the SpectroTranslator
metallicity (fehpred_5)
342-363 E22.19 --- [Fe/H]16 16th percentile of the SpectroTranslator
metallicity (fehpred_16)
365-386 E22.19 --- [Fe/H]50 50th percentile of the SpectroTranslator
metallicity (fehpred_50)
388-409 E22.19 --- [Fe/H]84 84th percentile of the SpectroTranslator
metallicity (fehpred_84)
411-432 E22.19 --- [Fe/H]95 95th percentile of the SpectroTranslator
metallicity (fehpred_95)
434-454 E21.19 --- e_[Fe/H]50 Systematic error of the SpectroTranslator on
the metallicity (fehpred_ERR)
456-477 E22.19 --- [Mg/Fe]5 5th percentile of the SpectroTranslator
magnesium over iron ratio (Mgfepred_5)
479-500 E22.19 --- [Mg/Fe]16 16th percentile of the SpectroTranslator
magnesium over iron ratio (Mgfepred_16)
502-523 E22.19 --- [Mg/Fe]50 50th percentile of the SpectroTranslator
magnesium over iron ratio (Mgfepred_50)
525-546 E22.19 --- [Mg/Fe]84 84th percentile of the SpectroTranslator
magnesium over iron ratio (Mgfepred_84)
548-569 E22.19 --- [Mg/Fe]95 95th percentile of the SpectroTranslator
magnesium over iron ratio (Mgfepred_95)
571-591 E21.19 --- e_[Mg/Fe]50 Systematic error of the SpectroTranslator on
the magnesium over iron ratio
(Mgfepred_ERR)
593-614 E22.19 km/s HV5 5th percentile of the SpectroTranslator
heliocentric velocity (VHELIOpred5)
616-637 E22.19 km/s HV16 16th percentile of the SpectroTranslator
heliocentric velocity (VHELIOpred16)
639-660 E22.19 km/s HV50 50th percentile of the SpectroTranslator
heliocentric velocity (VHELIOpred50)
662-683 E22.19 km/s HV84 84th percentile of the SpectroTranslator
heliocentric velocity (VHELIOpred84)
685-706 E22.19 km/s HV95 95th percentile of the SpectroTranslator
heliocentric velocity (VHELIOpred95)
708-728 E21.19 km/s e_HV50 Systematic error of the SpectroTranslator on
the heliocentric velocity (VHELIOpredERR)
730-734 A5 --- Flagmii [false/true ] If True, indicates that an
input was missing for the intrinsic network
(Flagmissinginputs_intrinsic)
736-740 A5 --- Flagmie [false/true ] If True, indicates that an
input was missing for the extrinsic network
(Flagmissinginputs_extrinsic)
742-901 A160 --- Com Explication of why at least one of the
quality flag was set on False
(Qflag_comments)
903-907 A5 --- QflagIi [false/true ] Quality flag that indicates
that the all the input of the intrinsic
network are inside the parameter space
covered by the training sample
(QflagInputintrinsic)
909-913 A5 --- QflagOi [false/true ] Quality flag that indicates
that the all the Output of the intrinsic
network are inside the parameter space
covered by the training sample
(QflagOutputintrinsic)
915-919 A5 --- QflagIe [false/true ] Quality flag that indicates
that the all the input of the extrinsic
network are inside the parameter space
covered by the training sample
(QflagInputextrinsic)
921-924 A4 --- QflagOe [true] Quality flag that indicates that the
all the Output of the extrinsic network are
inside the parameter space covered by the
training sample (QflagOutputextrinsic)
--------------------------------------------------------------------------------
Byte-by-byte Description of file: a2g.dat
--------------------------------------------------------------------------------
Bytes Format Units Label Explanations
--------------------------------------------------------------------------------
1- 19 A19 --- APOGEE APOGEE observation identifier,
2MHHMMSSss+DDMMSSs) (APOGEE_ID)
21- 39 I19 --- GaiaDR3 Gaia DR3 unique source identifier
(source_id)
41- 60 F20.13 K Teff5 5th percentile of the SpectroTranslator
effective temperature (TEFFpred5)
62- 81 F20.13 K Teff16 16th percentile of the SpectroTranslator
effective temperature (TEFFpred16)
83-102 F20.13 K Teff50 50th percentile of the SpectroTranslator
effective temperature (TEFFpred50)
104-123 F20.13 K Teff84 84th percentile of the SpectroTranslator
effective temperature (TEFFpred84)
125-144 F20.13 K Teff95 95th percentile of the SpectroTranslator
effective temperature (TEFFpred95)
146-169 F24.18 K e_Teff50 Systematic error of the SpectroTranslator on
the transformation of the effective
temperature (TEFFpredERR)
171-192 E22.19 [cm/s2] logg5 5th percentile of the SpectroTranslator
surface gravity (LOGGpred5)
194-215 E22.19 [cm/s2] logg16 16th percentile of the SpectroTranslator
surface gravity (LOGGpred16)
217-238 E22.19 [cm/s2] logg50 50th percentile of the SpectroTranslator
surface gravity (LOGGpred50)
240-261 E22.19 [cm/s2] logg84 84th percentile of the SpectroTranslator
surface gravity (LOGGpred84)
263-284 E22.19 [cm/s2] logg95 95th percentile of the SpectroTranslator
surface gravity (LOGGpred95)
286-306 E21.18 [cm/s2] e_logg50 Systematic error of the SpectroTranslator on
the surface gravity (LOGGpredERR)
308-329 E22.19 --- [M/H]5 5th percentile of the SpectroTranslator
metallicity (MHpred_5)
331-352 E22.19 --- [M/H]16 16th percentile of the SpectroTranslator
metallicity (MHpred_16)
354-375 E22.19 --- [M/H]50 50th percentile of the SpectroTranslator
metallicity (MHpred_50)
377-398 E22.19 --- [M/H]84 84th percentile of the SpectroTranslator
metallicity (MHpred_84)
400-421 E22.19 --- [M/H]95 95th percentile of the SpectroTranslator
metallicity (MHpred_95)
423-443 E21.18 --- e_[M/H]50 Systematic error of the SpectroTranslator on
the metallicity (MHpred_ERR)
445-466 E22.19 --- [Mg/Fe]5 5th percentile of the SpectroTranslator
magnesium over iron ratio (MGFEpred_5)
468-489 E22.19 --- [Mg/Fe]16 16th percentile of the SpectroTranslator
magnesium over iron ratio (MGFEpred_16)
491-512 E22.19 --- [Mg/Fe]50 50th percentile of the SpectroTranslator
magnesium over iron ratio (MGFEpred_50)
514-535 E22.19 --- [Mg/Fe]84 84th percentile of the SpectroTranslator
magnesium over iron ratio (MGFEpred_84)
537-558 E22.19 --- [Mg/Fe]95 95th percentile of the SpectroTranslator
magnesium over iron ratio (MGFEpred_95)
560-580 E21.18 --- e_[Mg/Fe]50 Systematic error of the SpectroTranslator on
the magnesium over iron ratio
(MGFEpred_ERR)
582-603 E22.19 km/s HV5 5th percentile of the SpectroTranslator
heliocentric velocity (VHELIOpred5)
605-626 E22.19 km/s HV16 16th percentile of the SpectroTranslator
heliocentric velocity (VHELIOpred16)
628-648 E21.18 km/s HV50 50th percentile of the SpectroTranslator
heliocentric velocity (VHELIOpred50)
650-671 E22.19 km/s HV84 84th percentile of the SpectroTranslator
heliocentric velocity (VHELIOpred84)
673-694 E22.19 km/s HV95 95th percentile of the SpectroTranslator
heliocentric velocity (VHELIOpred95)
696-716 E21.18 km/s e_HV50 Systematic error of the SpectroTranslator
on the heliocentric velocity
(VHELIOpredERR)
718-722 A5 --- Flagmii [false/true ] If True, indicates that an
input was missing for the intrinsic network
(Flagmissinginputs_intrinsic)
724-728 A5 --- Flagmie [false/true ] If True, indicates that an
input was missing for the extrinsic network
(Flagmissinginputs_extrinsic)
730-890 A161 --- Com Explication of why at least one of the
quality flag was set on False
(Qflag_comments)
892-896 A5 --- QflagIi [false/true ] Quality flag that indicates
that the all the input of the intrinsic
network are inside the parameter space
covered by the training sample
(QflagInputintrinsic)
898-902 A5 --- QflagOi [false/true ] Quality flag that indicates
that the all the Output of the intrinsic
network are inside the parameter space
covered by the training sample
(QflagOutputintrinsic)
904-908 A5 --- QflagIe [false/true ] Quality flag that indicates
that the all the input of the extrinsic
network are inside the parameter space
covered by the training sample
(QflagInputextrinsic)
910-913 A4 --- QflagOe [true] Quality flag that indicates that the
all the Output of the extrinsic network are
inside the parameter space covered by the
training sample (QflagOutputextrinsic)
--------------------------------------------------------------------------------
Acknowledgements:
Guillaume Thomas, gthomas(at)iac.es
(End) Patricia Vannier [CDS] 19-Aug-2024