Extinction

SLUG can apply extinction to a calculated spectrum. The extincted spectrum is computed as

\[L_{\lambda,\mathrm{ext}} = L_\lambda \exp\left(-A_V C_{\lambda,\mathrm{ext}}\right),\]

where \(L_{\lambda,\mathrm{ext}}\) is the extincted spectrum, \(L_\lambda\) is the intrinsic spectrum, \(A_V\) is the V-band extinction, and \(C_{\lambda,\mathrm{ext}}\) is a dimensionless extinction curve. The extinction curve is normalized so that when \(A_V = 1\)

\[-2.5 \log_{10}\frac{\langle L_{\lambda,\mathrm{ext}} \rangle_V}{\langle L_{\lambda} \rangle_V} = 1\,\mathrm{mag},\]

where \(\langle\cdot\rangle_V\) is the spectrum averaged over a Johnson V filter; see Photometry and Filters for the precise definition of the filter average.

When extinction is enabled (by setting values of \(A_V\) and the extinction curve) as described below, the output will contain an extincted spectrum as well as the intrinsic spectrum. Note that the extincted spectrum may be truncated to a wavelength grid that is narrower than the one used for the intrinsic spectrum if the extinction curve selected does not cover the full wavelength range of the intrinsic spectrum.

If nebular emission is enabled (see Nebular Emission), it will also contain spectra and photometry computed on the nebular-reprocessed spectrum. Thus in the most general case when both nebular reprocessing and extinction are enabled, SLUG will output four versions of each spectrum: the intrinsic stellar spectrum, the spectrum after nebular processing, the stellar spectrum with extinction applied but not nebular processing, and the spectrum processed by the nebula and then with extinction applied to it. If photometry is enabled (see Photometry and Filters), each output spectrum will also have a corresponding set of photometric outputs computed. See Output Files and Format for an explanation of how these outputs are stored in the output file, and slugpy for an explanation of how to access them using the slugpy reader.

Setting the Extinction

The values of \(A_V\) used in the simulations are controlled by keywords in the Extinction Keywords section of The SLUG Input Deck. The keyword AV gives the extinction (or distribution of extinctions) for stars that are part of clusters, while in galaxy simulations AV_field provides a separate distribution of extinctions that is applied only to “field” stars that are not part of clusters.

Finally, SLUG allows nebular emission to be subjected to a different amount of extinction than stellar emission, consistent with observations showing that the nebular emission lines are usually more heavily extincted than the stellar continuum (e.g., Calzetti et al. 1994). The keyword extinct.neb_factor (see Extinction Keywords) specifies the distribution of the ratio of nebular extinction to stellar extinction, \(A_{V,\mathrm{neb}}/A_{V,\mathrm{star}}\). This factor is applied to both clusters whose extinction is set by AV and field stars whose extinction is set by AV_field. For example, if extinct.neb_factor is set to 2, then nebular emission will be extincted by twice the \(A_V\) that is applied to stellar emission. Note that neb_factor can be a PDF, so that the ratio can be drawn from a distribution rather than being a fixed number.

Extinction Curves

The model keyword in the Extinction Keywords specifies the shape of the extinction curve to apply. Available extinction curves are registered in a TOML registry file (defaults to data/extinct/extinct.toml, but this can be overridden by setting the registry keyword). The curves included in the default data set are:

Bouchet_SMC: the Bouchet SMC extinction curve (Bouchet et al. 1985).
Calzetti_LMC: the Calzetti LMC extinction curve (Landini et al. 1984; Fitzpatrick 1999; D. Calzetti, personal communication).
Calzetti_MW: the Calzetti Milky Way extinction curve (Landini et al. 1984; Fitzpatrick 1999; D. Calzetti, personal communication).
Calzetti_starburst: the Calzetti starburst attenuation curve (Calzetti et al. 2000).
Draine_MW_RV3.1: the Draine (2003) Milky Way extinction curve, for a total-to-selective extinction ratio \(R_V = 3.1\) (Weingartner & Draine 2001; Li & Draine 2001; Draine 2003a,b,c).
Draine_MW_RV4.0: the same Draine (2003) Milky Way extinction curve, for \(R_V = 4.0\).
Draine_MW_RV5.5: the same Draine (2003) Milky Way extinction curve, for \(R_V = 5.5\).

The full list of references for each curve, which should be cited in any publication using it, is provided in the data/extinct/extinct.toml registry.

Adding New Extinction Curves

Users can add new extinction curves by adding to the default HDF5 file (data/extinct/extinct.h5) and adding a corresponding registry entry in the registry file, or by creating a new HDF5 and a corresponding registry.

The required format for the HDF5 file is as follows.

An extinction curve file contains one top-level HDF5 group per curve, named exactly as the curve is named in the curves list of the registry entry described below (e.g. a group named Calzetti_starburst). Each such group must contain two 1D datasets of equal length:

wavelength: the wavelength grid, in Angstrom, on which the curve is tabulated, sorted in order of increasing wavelength. A star (or, for a continuously-distributed population, a wavelength point) that falls outside this grid’s own coverage is simply dropped from the extincted spectrum, rather than extrapolated – see the note on truncation above.
kappa: the extinction curve itself, \(C_{\lambda,\mathrm{ext}}\), at each wavelength in wavelength. Only the curve’s overall shape matters: SLUG automatically rescales whatever is provided here so that it corresponds to \(A_V = 1\) mag, as described above, so kappa may be tabulated in any convenient, even arbitrary, units or overall normalization.

The required format for the registry entry is as follows.

An extinction registry is a TOML file with a top-level file key (a string): the path to the HDF5 file described above, resolved relative to the directory containing the registry file itself, exactly as for a track, spectra, or filter registry’s own file key (see Standard Track Sets). It must also have a top-level curves key (an array of strings), listing every curve name the registry describes; the default registry, data/extinct/extinct.toml, lists the standard curves described above.

For every name appearing in curves, the registry may (but is not required to) also contain a top-level table of that same name (e.g. [Calzetti_starburst]). SLUG itself never reads anything from this table – a curve’s actual data always come from the HDF5 file – but the following keys are included in the standard registry, and recommended for any new curve added in the same way, for documentation and citation purposes:

reference: One literature reference (as a string) or several (as an array of strings) that should be cited by anyone using this curve.
reference_url: One URL string, or an array of them matching reference, giving a link (e.g. an ADS abstract page) for each reference. A reference with no readily citable URL (e.g. a personal communication) may simply be omitted here.
description: A short, human-readable description of the curve.

The scripts used to generate the existing extinction curve data set are provided in data/tools/extinct, and these can serve as a guide.