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 is a summary of the method used in Williams et al. 2018 to assign SEDs to each galaxy.
First, they assign galaxy masses and type (star-forming/quiescent). For the star-forming galaxies, they also assign absolute UV magnitude and UV continuum slope.
By construction, the galaxies follow known \(M_{UV}\)–\(M_*\) and \(\beta\)–\(M_{UV}\) trends.
The goal is to assign SEDs that are consistent with these galaxy properties.
They used BEAGLE to create a parent catalog; this is a tool for modeling and interperating spectrophotometric galaxy SEDs. Self-consistent approach that describes stellar emission and its transfer through the ISM and IGM.
If available, they used BEAGLE to fit SED models to known observations; I will skip this step.
When the properties extend beyond the current measurements of real sources, they use BEAGLE to produce theoretical SEDs covering a range of parameters that can be matched to the redshift, stellar mass and integrated properties of the galaxies.
They altered:
\(z\): redshift
\(M_*\) stellar mass
\(\tau\): time scale of stellar mass formation
\(\hat{\tau}_V\): \(V\) band attenuation optical depth
\(\log U_S\): effective gas ionization parameter
\(Z\): metallicity
\(t\): age of oldest stars in the galaxy
Section 3.4.3 outlines how they constrain these parameters, to make sure they have reasonable SEDs. From my understanding, they create a grid in mass and redshift; based on the mass–metallicity relationship, they assign a range of realistic metallicities. Then based on the mass–metallicity–star formation history relation, they assign realistic SFHs, ect.
For quiescent galaxies (section 4.2), they do not need to assign all of these properties; they only assign \(t\), \(\tau\) and \(Z\).
This creates a grid of SEDs.
They match the galaxy to the closest SED in terms of \(z\), \(M_*\), \(M_{UV}\) and \(\beta\) (its not clear to me how thet determine which is the “closest”).
They then shift the redshift of the SED to match.