J/ApJS/246/9    Stellar parameters of LAMOST stars using SLAM    (Zhang+, 2020)

Deriving the stellar labels of LAMOST spectra with the Stellar LAbel Machine (SLAM). Zhang Bo, Liu C., Deng L.-C. <Astrophys. J. Suppl. Ser., 246, 9 (2020)> =2020ApJS..246....9Z 2020ApJS..246....9Z
ADC_Keywords: Abundances; Spectra, optical; Models; Stars, giant; Surveys Keywords: Astronomical methods ; Astronomy data analysis ; Bayesian statistics ; Stellar abundances ; Chemical abundances ; Fundamental parameters of stars ; Catalogs ; Surveys Abstract: The LAMOST survey has provided 9 million spectra in its Data Release 5 (DR5) at R∼1800. Extracting precise stellar labels is crucial for such a large sample. In this paper, we report the implementation of the Stellar LAbel Machine (SLAM), which is a data-driven method based on support vector regression (SVR), a robust nonlinear regression technique. Thanks to the capability to model highly nonlinear problems with SVR, SLAM can generally derive stellar labels over a wide range of spectral types. This gives it a unique capability compared to other popular data-driven methods. To illustrate this capability, we test the performance of SLAM on stars ranging from Teff∼4000 to ∼8000K trained on LAMOST spectra and stellar labels. At g-band signal-to-noise ratio (S/Ng) higher than 100, the random uncertainties of Teff, logg, and [Fe/H] are 50K, 0.09dex, and 0.07dex, respectively. We then set up another SLAM model trained by APOGEE and LAMOST common stars to demonstrate its capability of dealing with high dimensional problems. The spectra are from LAMOST DR5 and the stellar labels of the training set are from APOGEE DR15, including Teff, logg, [M/H], [α/M], [C/M], and [N/M]. The cross-validated scatters at S/Ng∼100 are 49K, 0.10dex, 0.037dex, 0.026dex, 0.058dex, and 0.106dex for these parameters, respectively. This performance is at the same level as other up-to-date data-driven models. File Summary: -------------------------------------------------------------------------------- FileName Lrecl Records Explanations -------------------------------------------------------------------------------- ReadMe 80 . This file catalog.dat 166 5378550 The Catalog of the 5 million converged LAMOST DR5 K giant stars with SLAM-derived stellar labels -------------------------------------------------------------------------------- See also: V/153 : LAMOST DR4 catalogs (Luo+, 2018) V/164 : LAMOST DR5 catalogs (Luo+, 2019) J/other/RAA/11.924 : Atmospheric parameters for 771 stars (Wu+, 2011) J/ApJ/808/16 : Cannon: a new approach to determine abundances (Ness+, 2015) J/AJ/151/144 : ASPCAP weights for APOGEE chemical elements (Garcia+, 2016) J/ApJ/836/5 : Abundances of LAMOST giants from APOGEE DR12 (Ho+, 2017) J/MNRAS/478/4513 : GALAH Survey DR2 (Buder+, 2018) J/ApJS/241/32 : OB stars from the LAMOST DR5 spectra (Liu+, 2019) J/ApJS/244/8 : M-type stars in LAMOST DR5 (Zhong+, 2019) http://www.sdss.org/ : SDSS home page Byte-by-byte Description of file: catalog.dat -------------------------------------------------------------------------------- Bytes Format Units Label Explanations -------------------------------------------------------------------------------- 1- 9 I9 --- ObsID [101008/585216248]?=- LAMOST identifier 11- 22 F12.8 deg RAdeg Right ascension (J2000) (ra) 24- 34 F11.8 deg DEdeg Declination (J2000) (dec) 36- 45 F10.6 --- snrg [0/998]?=- g-band signal-to-noise ratio of the spectra 47- 53 F7.1 K Teff [-7492/15227] SLAM-predicted effective temperature (teff_slam) 55- 61 F7.3 [cm/s2] logg [-41.7/31.3] SLAM-predicted log of surface gravity (logg_slam) 63- 70 F8.4 [-] [M/H] [-21.6/12] SLAM-predicted metallicity (mh_slam) 72- 78 F7.4 [-] [a/M] [-4.2/7.1] SLAM-predicted [α/M] (am_slam) 80- 86 F7.4 [-] [C/M] [-9/7.5] SLAM-predicted [C/M] (cm_slam) 88- 94 F7.4 [-] [N/M] [-9.5/15.5] SLAM-predicted [N/M] (nm_slam) 96-102 F7.1 K e_Teff [38.8/241.3] Teff uncertainty (tefferrslam) 104-109 F6.4 [cm/s2] e_logg [0.06/0.7] logg uncertainty (loggerrslam) 111-116 F6.4 [-] e_[M/H] [0.02/0.5] Uncertainty in [M/H] (mherrslam) 118-123 F6.4 [-] e_[a/M] [0.01/0.2] Uncertainty in [a/M] (amerrslam) 125-130 F6.4 [-] e_[C/M] [0.04/0.2] Uncertainty in [C/M] (cmerrslam) 132-137 F6.4 [-] e_[N/M] [0.07/0.3] Uncertainty in [N/M] (nmerrslam) 139-149 F11.4 --- O-C [0.005/243963] rms deviation between the observed and fitted spectra (rmseppix_slam) 151 I1 --- K [0/1] K giant star selected (1=yes) (ind_kgiant) 153-164 F12.6 --- D [0.0005/16038]?=- Distance D (DHo17) (1) 166 I1 --- Fl [0/1] APOGEE observation flag (1:observed by APOGEE) (ind_common) -------------------------------------------------------------------------------- Note (1): We calculated the "label-distance" D defined by Ho+ (2017, J/ApJ/836/5), i.e., Equation (11): D=(Teff,SLAM-Teff,LAMOST)2/(100K)2 +(loggSLAM-loggLAMOST)2/(0.2dex)2 +([M/H]SLAM-[Fe/H]LAMOST)2/(0.1dex)2 To be consistent with Ho+ 2017, we consider stars with D<2.5 as K giant stars. -------------------------------------------------------------------------------- History: From electronic version of the journal
(End) Emmanuelle Perret [CDS] 01-Apr-2020
The document above follows the rules of the Standard Description for Astronomical Catalogues; from this documentation it is possible to generate f77 program to load files into arrays or line by line