BEAST Grid Inputs¶
Below are details regarding grid choices for stellar evolution models, spectral libraries, and dust extinction laws.
Stellar Evolution Models¶
Stellar evolution models provide the BEAST model grid with stellar parameters (L, Teff, log g) as a function of age, metallicity, and stellar mass. The BEAST is currently implemented to obtain this information from stellar isochrone sets. These isochrones are obtained at run-time via webforms that yield interpolated data products and stored as CSV files.
The Padova family of stellar evolution models spans multiple generations of work (Bertelli+94, Girardi+00, Girardi+02; Marigo+08, Girardi+10), including the most recent set of PARSEC and COLIBRI models (Bressan+12, Marigo+17).
PARSEC models accessed via CMD webform by
modeltype(default=’parsec12s_r14’ for PARSEC+COLIBRI): isochrone model set - alternate choices=[‘parsec12s’ for PARSEC 1.2S, ‘2010’ for Marigo08+Girardi10];
filterPMS(default=False): remove pre-main sequence stars (M < 12 Msun).
Details = age: 0 to 13.5 Gyr; metallicity: [M/H] from -2.2 to +0.5 (0.0001≤Z≤0.02; 122 pts), also super-solar available (Z = 0.03,0.04,0.06); solar abundance: Z_sun=0.0152 based on Caffau+11.
The MIST models (Choi+16) are computed using MESA (Paxton+11,13,15).
MIST Isochrone Interpolator accessed via webform by
rotation(default=’vvcrit0.0’): select rotating or non-rotating models - alternate choice=[‘vvcrit0.4’ for initial rotation of 0.4 x critical rotation].
Details = age: log(Age) from 5.0 to 10.3 (default sample = 0.05 dex); metallicity: [Fe/H] from -4 to +0.5 (default sample = 0.25 dex), all solar scaled in Version 1.2; solar abundance: Z_sun=0.0142 based on Asplund+09 (Z_proto=0.0142, Z_photo=0.0134).
Stellar atmosphere models provide the BEAST model grid with stellar spectra (surface flux as function of wavelength) as a function of Teff, log g, and metallicity. The BEAST currently uses theoretical spectral libraries stored as static libraries. Model files store spectra in flux format with units of erg/s/cm2/AA.
ATLAS9 model atmospheres by Castelli & Kurucz (2004; CK04), the industry standard for LTE stellar atmospheres that spans a wide range of L and Teff parameter space and a broad range of wavelengths (0.10-10.0 micron) at low spectral resolution (20 Ang/pix; 2x sampling of R~100 at 4000 Ang). Z grid spans -2.5 to +0.5 at 0.5 dex sampling. Models are computed using solar abundances from Grevesse & Sauval (1998; Z=0.0169).
Non-LTE hot star (Teff > 15,000 K) atmosphere models (OSTAR and BSTAR) by Lanz & Hubeny (2003, 2007), using spectra computed for Cloudy by Peter van Hoof at R~900 (sampled at R~1800). Z grid spans 5 pts from 1/10-2x Solar, plus additional 4 pts from 1/1000-1/30x Solar for OSTAR grid. Models are computed using solar abundances from Grevesse & Sauval (1998; Z=0.0169).
PHOENIX model atmospheres by Allard et al. 2016, specializing in cool star atmospheres (Teff < 6000 K). Intrinsically high-res, resampled to 2 Ang/pix grid (medres) and to match CK04 grid (lores). Models are computed using solar abundances from Asplund et al. 2009 (Z=0.0134).
Adjustable Parameter =
medres(default=True): 2 Ang/pix resolution, or False for CK04 matched wavelength grid.
Future Addition – ATLAS9 model atmospheres computed by Bohlin et al. 2017 providing enhancement to CK04 in terms of spectral resolution, wavelength coverage, grid density.
ATLAS9 model atmospheres available at higher spectral resolution than CK04, but over limited wavelength range.
Recommendations: Tlusty+Kurucz as default, providing non-LTE models at high temperatures and standard ATLAS9 models elsewhere. BTSettl library provides improvement at low temperatures, while BOSZ (coming soon) or Munari provide higher spectral resolution if required.
Dust extinction is applied to model spectra before bandpass convolution, providing fully self-consistent treatment of reddening.
- Generalized_RvFaLaw (recommended)
allows for any choice of the extinction curve model for the A and B components
A: F19 - based on spectroscopy in UV and optical and does a bit better than F99 in the optical
B: G03_SMCBar - best average for the SMC “bumpless” extinction curve
In beast_settings file: extLaw = extinction.Generalized_RvFALaw(ALaw=extinction.Generalized_DustExt(curve=’F19’), BLaw=extinction.Generalized_DustExt(curve=’G03_SMCBar’))
- dust_extinction Package Extinction Curves (recommended)
extinction.Generalized_DustExt(): Wrapper for any extinction curve class available via dust_extinction python package.
Select curve class via string parameter:
Example call, for Fitzpatrick+19:
For the most possible models, see dust_extinction docs
beast.physicsmodel.dust.extinction_extension(recommended of sub 912 A extinction needed)
all the observation based models stop at or before 912 A as hydrogen dominates extinction below this wavelength
For work that interested in sub 912 A information (e.g., ionizing photon measurements), these models extend the dust extinction to 100 A by smoothly merging dust grain models with an observed extinction model starting at 1675 A. Thus, the extinction between around 1150 A and 1675 A is a combination of the observed extinction curve and the dust grain model extinction curve. Most observational based extinction curves do not go all the way to 912 A due to lack of measurements.
F19_D03_extension: Extension of Fitzpatrick+19 Milky Way Rv dependent model with Draine03 grain models.
G03_SMCBar_WD01_extension: Extension of Gordon+03 SMCBar average with the Weingarter & Draine01 SMCBar grain model.
- Gordon+16 (original mixture dust extinction model)
extinction.Gordon16_RvFALaw(): Mixture model of MW (Type A; Fitzpatrick 99) and SMC (Type B; Gordon+03) extinction curves.
Adjustable parameters include: A_V, R_V, and f_A.
- Fitzpatrick 99 (may be deprecated, superceded by dust_extinction package)
extinction.Fitzpatrick99(): Default Milky Way extinction curve model.
Adjustable parameters: R_V
- Gordon+03 (may be deprecated, superceded by dust_extinction package)
extinction.Gordon03_SMCBar(): Empirically-derived SMC Bar dust extinction curve.
Adjustable A_V, R_V fixed at 2.74
- Cardelli, Clayton, and Mathis 89 (may be deprecated, superceded by dust_extinction package)
extinction.Cardelli89(): Well-known Milky Way extinction curve model, but advise use of
dust_extinction F19model instead.
Adjustable parameters: R_V