Nicole Drakos
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
Luminosity Function
In this post, I showed some preliminary predictions for the galaxy catalog.
The luminosity function (LF) didn’t look quite right, and I’m exploring that a bit more in this post.
Results
Here I have plotted it again with the Bouwens et al. 2015 luminosity functions:
Update 10/02: The Bouwens relations should have only been for z>4… see next post
\(M_{\rm UV}\) assignment
First, I want to check if this is due to problems in the \(M_{\rm UV}\)–\(M\) relation. Here are how our current mock catalog relations compare to the relations in W18:
The \(M_{\rm UV}\)–\(M\) relation isn’t too different, but there are some discrepancies. If instead I draw \(M_{\rm UV}\) properties from the W18 relations, rather than get from our SEDs, as in this post, I can plot the LF and get
This looks a bit better, but it still doesn’t match, so I don’t think our \(M_{\rm UV}\) assignment is the problem.
Galaxy Completeness
The other possible problem is that the it is because the SMFs are not complete below \(\approx 10^9\,M_{\odot}\) (e.g. see this post).
To test the effect of missing low mass galaxies, I again used \(M_{\rm UV}\) values drawn from the W18 distribution, but I didn’t include galaxies with masses below \(10^{10}\,M_{\odot}\).
This makes a very large difference on the SMF, which makes me thing this is the reason we can’t reproduce the observational LFs.
For the presentation
Ideally I would have an estimate for how well we would be able to constrain the faint end of the luminosity function, but right now I don’t have enough low mass galaxies to resolved in the simulation, and I don’t think I have enough time to analyze the high resolution simulations before the conference.
Since I don’t need spatial information from this plot, I can just generate galaxy mass/redshifts from the distributions down to lower mass resolutions, and then run them through my SED assignment code and then plot the LF.
I think W18 used a mass limit of \(M_{\rm min} = 10^6\), and they were able to reproduce the LFs, so that should be enough. However, I only expect the SMFs for the \(2048^3\) simulations to be complete to \(\approx 10^7\,M_{\odot}\), so I will start with that mass limit. If there are issues reproducing the LFs with \(M_{\rm min}=10^7\,M_{\odot}\), we might need to consider adding subhalos analytically to the catalog.