Institute of Cosmology and Gravitation

MaStar
Stellar Population Models

12/06/2020: UPDATE to version v0.2

The MaStar Stellar Population Models (Maraston et al 2020) use the new SDSS-based stellar library MaStar (Yan et al 2019) to predict the spectral energy distributions of stellar population models with various chemical composition and age.

MaStar is an empirical library containing around 9,000 stellar spectra acquired with the SDSS 2.5m telescope, hence spectral resolution and wavelength range are identical to SDSS data.  The wavelength range is 3,600 to 10,300 A with a wavelength-dependent spectral resolution ranging from R=1,400 to R=2,200. The resolution array is provided with the model tables (see below). Wavelengths are in vacuum.

The models are based on the Maraston (2005) and Maraston & Strömbäck (2011) synthesis codes and input physics.

There are two model flavours with different stellar parameters for the MaStar sample, E-MaStar and Th-MaStar. In E-MaStar, stellar parameters are derived from fitting empirical stellar spectra from the MILES stellar library (Chen et al 2020, ApJ, submitted), while in Th-MaStar stellar parameters are derived from fitting stellar spectra of theoretical stellar atmospheres from MARCS and ATLAS (Hill et al 2020, in preparation).

The models are calculated for a wide range in age (t = 0.1 – 15 Gyr), metallicity ([Z/H] = -2.35 – 0.35 dex) and low-mass IMF slope (s = 0.3 – 3.8). These ranges are being extended through forthcoming releases thanks to the ongoing MaStar campaign. Please refer to the paper and in particular to Table 1 for a description of the confidence range of models.

Data access

The model data can be downloaded as gzipped fits file MaStar_SSP_v0.2.fits.gz (version v0.2). The models are stored in the third and fourth extensions as a 4-dimensional matrix. HDU3 contains the Th-MaStar models and HDU4 the E-MaStar models. In these, one dimension is model flux, the other three are the three model parameters age t, metallicity Z and low-mass IMF slope s. The equivalent 4-dimensional matrix of the model parameter arrays is stored in HDU1, while HDU2 contains the 1-dimensional wavelength array (linear in Angstrom, vacuum) and the array of spectral resolution R.

Python functions to read in the data are provided with script MaStar_SSP.py. The python script dial_MaStar_SSP.py plots a model spectrum for an input set of values for t, Z, x.

MaStar Stellar Parameters

Here we provide stellar parameters for MaStar spectra, namely: effective temperature, surface gravity, [Fe/H] (for all MaStar spectra, Hill et al. 2022a) and [alpha/Fe] (for a subset of MaStar spectra, Hill et al. 2022b (in press)). Parameters based on solar-scaled [alpha/Fe] models are here and variable [alpha/Fe] parameters are here.

Parameters were calculated by fitting theoretical spectra from model atmospheres (from MARCS (Gustafsson et al. 2008), add cite and BOSZ (Mészáros et al. 2012; Bohlin et al. 2017), to observed MaStar spectra using the penalised pixel-fitting method (pPXF), utilising the full wavelength range of MaStar (3,600 – 10,300 A). The parameters are sampled using an MCMC approach and utilise ancillary data from GAIA to break gravity degeneracies at the cold end of the HR diagrams.

Note: our parameters can be also downloaded from the MaStar value added catalogue, which additionally includes a calibrated median of ours (Hill et al. 2022a, Hill et al. 2022b (in press)) and other parameters calculated for MaStar spectra (Imig et al. 2021, Lazarz et al. 2022 (in press), Chen et al. (in prep)), where the calibration is performed in metallicity using APOGEE results. For the solar-scaled parameters, the ‘valid’ column may be used to filter parameters we deem as reliable based on the chi square of the model fit. The variable alpha catalogue provides two valid columns. The ‘valid’ column only applies to temperature, log g and [Fe/H] and uses the same chi square criteria as in the solar-scaled set, as well as additional convergence checks that are described in 2022b. The ‘valid_am’ column applies to the [alpha/Fe] column and uses the error on [alpha/Fe] to remove any remaining unreliable parameters. Further details can be found in the original papers and a description of the data columns found in the header of each file.

Carbon- and Oxygen-rich TP-AGB stars in MaStar

Here we provide the properties of Carbon and Oxygen-rich TP-AGB spectra we identified in MaStar (Hill et al. 2022c). The full table can be downloaded here.

Properties include: the MaStar identification, median signal-to-noise ratio per pixel, star coordinates, calculated colours in both SDSS and Johnson-Cousins systems, whether the spectrum is from the badspec catalogue (flag=1) or not (flag=0), the spectral type and the class given by full-spectral fitting with Lançon & Mouhcine 2002 or from the X-shooter (Gonneau et al. 2017) templates. The table available here contains parameters relating to all C- and O-types identified in the paper, with each row matching Figures A1 and A2 of Hill et al. 2022c, reading them from left to right and top to bottom.

The identification is performed with a variety of methods, based on optical SDSS filters u, g, r, i, z only and can be used for other surveys using similar photometry.

 Version Log

  • 12/06/2020, v0.2: Bug fix to correct spectral resolution, resolution array added, artefact removed from metal-poor Th-MaStar model
  • 14/11/2019, v0.1: First release based on SDSS/MaNGA data release DR16.

Twitter

Facebook