The Luminosity Function of Galaxies

Very Important Point

All data distributed here has been FULLY superceded by the pair of papers Blanton et al. (2003c) and Blanton et al. 2003d (for which I give web pages and data distributions).

Introduction

This page presents the non-parametric fits to the luminosity function of galaxies from commissioning data of the Sloan Digital Sky Survey. This survey will eventually obtain one million galaxy redshifts for galaxies with r<17.65 or so. This work is based only on one percent of that data. Most of the pertinent information regarding the measurement of the luminosity function is in the accompanying paper:

Blanton, M. R., et al., SDSS Collaboration 2001, ``The Luminosity Function of Galaxies in SDSS Commissioning Data,'' AJ, 121, 2358 (astro-ph/0012085) [Full Referreed Version]

For more general information about the SDSS, see the technical overview given by:

York, D. G., et al., SDSS Collaboration 2000, ``The Sloan Digital Sky Survey: Technical Summary,'' AJ, 120, 1579 [Full Referreed Version]

Description of the results is given below. Please contact me with problems, questions, discrepancies, etc.

Luminosity Function Fits

The non-parametric fits to the luminosity function in each band are given in this directory in files with the naming convention: lf.[band].[omegam].[omegal].dat, where "omegam" is the matter density and "omegal" is the contribution from the cosmological constant. Each line represents a bin, and has the format:

[M] [PhiHat(M)] [LogError]

"M" is the center of each absolute magnitude bin; "PhiHat(M)" is the number density per unit magnitude in that bin; "LogError" is the logarithmic (base 10) error in the number density (these are really the diagonal elements of the covariance matrix, which is -- it turns out -- non-diagonal). Values of zero indicate that no data is available for that bin.

Joint Luminosity-Property Fits

The non-parametric fit to the joint distribution of luminosity and some other property are given in the following files for each property:

Half-light Surface Brightness (Figure 10)
Petrosian Color (g-r) (Figure 13)
Inverse Concentration Index (1/c) (Figure 16)

These results are only given for a flat, lambda-dominated universe with a matter density of 0.3. The format is very similar to that of the luminosity function. Each line represents a bin, and has the format:

[M] [Property] [PhiHat(M,Property)] [LogError]

"M" is the center of each absolute magnitude bin; "Property" is the center of each bin in the given property; "PhiHat(M,Property)" is the number density per unit magnitude per unit property in that bin; "LogError" is the logarithmic (base 10) error in the number density (these are really the diagonal elements of the covariance matrix, which is -- it turns out -- non-diagonal). Values of zero indicate that no data is available for that bin.

Note that there was an error in these joint LF files previous to 3/29/2002 which was that they were normalized incorrectly. The correct normalization is now given in the tables.

Source Code

If you would like to rummage through the source code used to perform these calculations, feel free to do so in this directory. It is not very well commented (and not very fast) but I include it here on the off chance that someone wants to check the code directly. You won't be able to use it very easily, because you won't have the right input files (which I cannot give you --- it would amount to a data release which I am not authorized to do). Also, the outputs do not give things in exactly the right units; post-processing code is putting things in the right units at the moment. In the simple luminosity function case, this involves multiplying by 0.4 to make the function per unit magnitude rather than per base-10 logartithm.

Snail Mail: Michael Blanton; Email: blanton at nyu.edu