Project II: Selection Effects on Galaxy Samples as a Function of Redshift


When studying galaxy samples over a wide range in redshift, our "observing window" into the universe is largely determined by a selection function that includes, among other effects:

In this project, you will learn to estimate the limitations imposed by such effects in current studies of galaxies at various redshifts. The main goal is to determine what fraction and what type of galaxies in our nearby universe can be observed at higher redshifts within the current limitations of our astronomical instrumentation. In the process, you will gain an understanding of the main selection effects that affect most comparison studies between the low- and high-redshift universe aiming at unveiling the nature of distant galaxies, and assessing their evolution over cosmological timescales.

In summary, the project consists of plotting a representative sample of the galaxy population in the nearby universe (the so-called "reference sample") in a 3-D observational parameter space defined by:

and identify those galaxies that can be seen at a given redshift within the typical observational constrains of most galaxy surveys. Below you will find some guidelines on how to carry out this project. However, I emphasise that not all the required information is provided. You will have to: (i) do your own research into mining data from astronomical databases to draw the reference galaxy sample in the local universe; (ii) derive your own formulae to calculate the various parameters in the 3-D observational parameter space that will serve as the framework for this project; (iii) consult the literature for any applicable passband or color transformations; and (iv) use various codes, including your own, to calculate cosmological distances and identify the region of the parameter space that can be observed at a given redshift.


The reference sample will consist of 10,000 galaxies with redshifts z<=0.05 selected from the Sloan Digital Sky Survey. For each galaxy, you will need to extract the following information listed in the "SpecPhotoall" and "Galaxy" tables: identification number (objid), redshift (z), total r-band apparent magnitude (r, or dered_r), total u-band apparent magnitude (u, or dered_u), and half-light radius in the r-band (petroR50_r). The "dered" magnitudes have already been corrected for Galactic extinction.

An example of the sintaxis used to retrieve data from the SDSS First data Release database is provided below:

select top 20 sp.z, g.petroR50_r

from specPhotoall as sp, Galaxy as g


(sp.objid=g.objid) and

(sp.z between 0.0 and 0.05)

For the purposes of this project, galaxies in this reference sample can roughly be divided into three "morphological" types depending on their observed colors (e.g., Fukugita. Shimasaku, and Ichikawa, 1995, PASP 107,945):

Internal extinction corrections, k-corrections, and evolutionary corrections will depend on the galaxy type. To apply the extinction correction you can use the analytical extinction law for our Galaxy derived by Cardelli, Clayton, and Mathis (1989; ApJ 345, 245) assuming the diffuse ISM case (RV=3.1): A(V)=0.1 mag for the spheroidal population; A(V)=1.5 mag for the spiral population, and A(V)=0.5 mag for the starburst population. The k-correction can be derived using the task kcor in the COSMOPACK package within IRAF, assuming "E" case for the spheroidal population; "Sa" case for the spiral population, and "burst" case for the starburst population. Also, you can assume the SDSS u and r filters are similar to the Uj and r filter options considered in the task kcor. No evolutionary correction will be applied in this project.

From the tabulated parameters, you will be able to derive: LR, SBeR, and (u-r)0. This is the data set that will be used as the reference sample.

For this project, you will assume that the observational limitations of a typical galaxy survey are:

These two limits will determine the fraction of galaxies of each type in the reference sample that can be observed at various redshifts. For this project you will investigate the following cases: z=1.0 and 3.0 for the "benchmark" cosmological model.

In addition, you will repeat the same analysis assuming the observational limiting magnitude and surface brightness in I band derived from the deepest image so far: the Hubble Ultra Deep Field (UDF).


The results will be presented in a written report: