Research Blog
Welcome to my Research Blog.
This is mostly meant to document what I am working on for myself, and to communicate with my colleagues. It is likely filled with errors!
This project is maintained by ndrakos
This post is on choosing the parameters in FSPS.
There are a large number of possible parameters that can be changed. Here is the description for the python-FSPS parameters:
The redshift of each galaxy, zred is known.
The age of the galaxy (in Gyr), tage can be varied. I think this is the age since the start of star formation. This is something I will treat as a free parameter.
I have turned on add_stellar_remnants, which includes stellar remnants in the mass composition, and turned off add_igm_absorption which includes IGM absorption via Madau (1995).
The options for the initial mass functions (IMFs) are Salpeter, Chabrier, Kroupa, van Dokkum, Dave or your own tabulated IMF. For now I will use a Chabrier 2003 IMF (imf_type=1), because that was what was used in Williams et al.
zcontinuous specifies how the interpolation in metallicity is done. I will use the option \(1\), in which the SSPs are interpolated to the value of logzsol before the spectra and magnitudes are computed.
The metallicity of each galaxy is then controlled by the parameter logzsol, which gives the metallicity in units of \(\log(Z/Z_\odot)\)
I’m going to use a delayed tau model.
The code has a delayed six parameter tau model (sfh=4), that has a tau model plus a constant component and a burst; I will leave the parameters that control the constant and burst components off.
The parameters are:
tau: e-folding time in Gyr. Range is \(0.1 <\tau <10^2\).
sf_start: start time of SFH, in Gyr. (is this the age of the universe? maybe I should just leave this to zero?)
sf_trunc: truncation time of SFH, in Gyr. Can I use this to control whether a galaxy is quiescent?
The dust absorption follows Charlot & Fall (2000). Dust emission follows Draine & Li (2007). I included the circumstellar dust models calibrated from DUSTY, but not the dust emission associated with an AGN torus.
I will use dust_type=2 for the extinction curve for now, as it seems to be the simplest. This uses the Calzetti et al. (2000) attenuation curve, where dust attenuation is applied to all starlight equally (not split by age), and therefore the only relevant parameter is dust2, which sets the overall normalization (but you must also set dust1=0.0 for this to work correctly).
I will turn on the nebular emission model (add_neb_emission=True)
(also include the nebular continuum, which it will do by default).
The nebular emission is controleld by the gas ionization parameter gas_logu, and the gas metallicity, gas_logz. As in Williams et al., I will approxiate the stellar and ISM metallicities as a single metallicity value (\(Z_{ISM} = Z\))
From this post, the parameters that will vary between each galaxy are:
zred – redshift of galaxy. This can be set
tage – age of galaxy in Gyr. I think this is probably the age since star formation begins. (Edit: this is actually the age of the universe… )
logzsol – metallicity
tau – e-folding time in Gyr for star formation
sf_start – start time of SFH, in Gyr
sf_trunc – end time of SFH, in Gyr
dust2 – dust attenuation
gas_logu – gas ionization parameter
gas_logz – gas metallicity
Next, I will look into how these effect the properties I want to match (e.g. mass, UV magnitude, UV continuum slope)