Stellar Parameters of BOSS M dwarfs in SDSS-V DR19

TL;DR

The paper presents SLAM, a data-driven SVR-based pipeline to derive metallicity, effective temperature, and surface gravity for SDSS-V BOSS M dwarfs from low-resolution optical spectra. By calibrating $[\mathrm{Fe/H}]$ with FGK+dwarf companions in wide binaries and $T_{\rm eff}$ and $\log g$ with APOGEE Net, SLAM delivers parameter estimates that agree with external benchmarks and quantify uncertainties that depend on spectral SNR. Validation on M+M binaries and cross-comparisons with Birky (2020), Behmard (2025), LAMOST, and Mann reveal small biases and robust consistency, while ASPCAP metallicities are corrected via a $T_{\rm eff,ASP}$ and $[\mathrm{Fe/H}]$-dependent calibration. The method’s applicability is bounded by the training domain $(\mathrm{Fe/H}, T_{\rm eff}, \log g)$ ranges, and the authors outline pathways to extend to metal-poor regimes with upcoming data and follow-up spectroscopy, enhancing metallicity studies for the Galactic M-dwarf population.

Abstract

We utilized the Stellar LAbel Machine (SLAM), a data-driven model based on Support Vector Regression, to derive stellar parameters ([Fe/H], $T_{\rm eff}$, and $\log{g}$) for SDSS-V M dwarfs using low-resolution optical spectra (R$\sim$2000) obtained with the BOSS spectrographs. These parameters are calibrated using LAMOST F, G or K dwarf companions ([Fe/H]), and APOGEE Net ($T_{\rm eff}$ and $\log{g}$), respectively. Comparisons of SLAM predicted [Fe/H] values between two components of M+M dwarfs wide binaries show no bias but with a scatter of 0.11 dex. Further comparisons with two other works, which also calibrated the [Fe/H] of M dwarfs by using the F/G/K companions, reveal biases of -0.06$\pm$0.16 dex and 0.02$\pm$0.14 dex, respectively. The SLAM-derived effective temperatures agree well with the temperature which is calibrated by using interferometric angular diameters (bias: -27$\pm$92 K) and those of the LAMOST (bias: -34$\pm$65 K), but are systematically lower than those from an empirical relationship between the color index and $T_{\rm eff}$ by 146$\pm$45 K. The SLAM surface gravity aligns well with those of LAMOST (bias: -0.01$\pm$0.07 dex) and those derived from the stellar mass and radius (bias: -0.04$\pm$0.09 dex). Finally, we investigated a bias in [Fe/H] between SLAM and APOGEE ASPCAP. It depends on ASPCAP's [Fe/H] and $T_{\rm eff}$, we provide an equation to correct the ASPCAP metallicities.

Figures (12)

• Figure 1: The color-absolute magnitude diagram ($BP\_RP_{0}$ vs. $M_{G0}$) of all stars classified as M-type stars by the BOSS spectroscopic reduction pipeline in DR19. The blue points represent objects that do not satisfy the criteria 1–4 defined in Section \ref{['sect:Identification of M dwarfs']}, and are therefore considered non-M dwarfs. Red points denote M dwarf candidates identified according to criteria 1–4.
• Figure 2: The top panel is the CMD of 1120 FGK+M wide binaries. The blue and red dots are the LAMOST FGK primaries and BOSS M secondaries, respectively. The bottom panel is the same as the top panel, but for the M+M wide binaries. The blue and red dots represent the M dwarf primaries and secondaries, respectively.
• Figure 3: The distributions of SLAM training labels, including $[\mathrm{Fe/H}]$, $T_{\mathrm{eff}}$ and $\log{g}$, which described in Section \ref{['sect:training data']}. The ranges of $[\mathrm{Fe/H}]$, $T_{\mathrm{eff}}$ and $\log{g}$ are (-1, 0.5) dex, (2900,4200) K and (4.4,5.1) dex, respectively.
• Figure 4: Top row: one-to-one comparisons between SLAM prediction values and the reference labels for $[\mathrm{Fe/H}]$ (FGK catalogue; left), $T_{\mathrm{eff}}$ (APOGEE Net; middle), and $\log{g}$ (APOGEE Net; right); points are colour–coded by BOSS spectral $\mathrm {SNR}$. Middle row: corresponding residuals, $\Delta[{\rm Fe/H}]_{\rm SLAM-FGK}$, $\Delta T_{\rm eff,SLAM-ApogeeNet}$, and $\Delta\log g_{\rm SLAM-ApogeeNet}$, as a function of the reference labels, with dashed lines marking zero offset. Bottom row: distributions of these residuals, with text boxes giving the median bias and standard deviation of difference (approximately $0.03\pm0.25$ dex in $[\mathrm{Fe/H}]$, $11\pm168$ K in $T_{\mathrm{eff}}$, and $0.00\pm0.10$ dex in $\log{g}$).
• Figure 5: The left panel shows the $\rm \Delta [Fe/H] (=[Fe/H]_{SLAM-FGK})$ versus the spectral $\mathrm {SNR}$ of the test dataset. Gray points are individual cross‐validation residuals for single stars. Stars are sorted by $\mathrm {SNR}$ and grouped into equal‐count $\mathrm {SNR}$ bins ($\sim$20); in each bin the blue circles with error bars show the median $\mathrm {CV\_bias}$ and its 1$\sigma$ uncertainty, while the red triangles with error bars show the median $\mathrm {CV\_scatter}$ and its uncertainty. The solid red curves are weighted fits to the binned scatters of the form $\sigma=a+b/\mathrm {SNR}^{c}$, with the best‐fitting relation and the valid $\mathrm {SNR}$ range indicated in the upper part of each panel. The middle and right panels are the same as the left panel, but for $T_{\mathrm{eff}}$ and $\log{g}$ , respectively.