Supplemental materials of
Clustering of High Redshift (z>2.9) Quasars from the Sloan Digital Sky Survey
high-resolution paper astro-ph Press Release

last updated: Feb 2007 by Yue Shen

Tables and Figures

datafile1.txt : the entile Table 1 in the paper.
fb1.eps : Fig. B1 of the paper; the illustrative figure for the SDSS tiling geometry.


The angular geometry description of the quasar clustering subsamle can be found in Appendix B of the paper. Files necessary to generate the masks are present here in fits format. You need to install the idlutils and idlspec2d products first. Use the READ_FITS_POLYGONS routine in idlutils to read these fits files.

The first two files are enough for your purpose of generating random catalogs in the same regions as the clustering subsample with or without bad imaging fields. The other three files are useful if you want to plot the targeting and tiling geometry as shown in Fig. B1 of the paper. Note that for the highz_spectro_window.fits file, the starting plate is 0716, because prior plates were associated with tiling chunks with targeting version lower than v3_1_0. However, a later plate number doesn't guarantee the plate is necessarily tiled in a tiling chunk with targeting version no lower than v3_1_0. In other words, you need to use both tilingChunks.fits (with targeting version incorporated) and highz_spectro_window.fits to reject lower target version regions.

: the set of balkanized spherical polygons describing the geometry of the clustering subsample. Note that here "balkanization" is only done within each tiling chunks (tileRun). Since different tiling chunks could overlap with each other (see the examples in Fig. B1 of the paper), you need to do "balkanization" again to get the fully "balknized" sectors, and to get the total net area covered by these sectors. Such a file is provided as full_sectors_balkans.fits, where the tiling chunk and plate info is gone.
bad_fields_balkans.fits : a secondary mask describing the regions of bad imaging fields.
targetingChunks.fits : the set of polygons describing the geometry of target chunks.
tilingChunks.fits : the set of polygons describing the geometry of tiling chunks.
highz_spectro_window.fits : the set of polygons describing the geometry of spectroscopic plates (tiles) starting from plate 0716.

How to determine whether or not a given point (xyz or RA/DEC) is in regions described by the set of polygons?
Use routine IS_IN_WINDOW in idlutils.

I also put the original files describing the boundaries of the targeting chunks and tiling chunks below (these files can also be obtained from CAS and/or DR5 site). There files are necessary if you want to generate the above fits files by yourself.

tilingBoundaries : table describing the boundaries of tiling chunks; its parent view is tilingGeometry.
targetingChunk : file describing the boundaries of targeting chunks (never used in generating full_sectors.fits, but is of general SDSS geometry interests).
maindr5spectro.par : file describing the geometry of spectroscopic plates.
runQA_DR5.csv (46M!!): file describing detailed SDSS run quality, used to identify bad imaging fields. After identifying bad imaging fields, you need to use the SDSS_RUN2WINDOW routine in the SDSS photoop product. (Yes, you have to install photoop product to use sdss_run2window and related routines).

The SDSS tiling geometry is rather complicated, please refer to the description in Appendix B of the paper for more information. If you REALLY want to play with spherical polygons for SDSS, you may want to check the SDSS glossary for some jargons and especially different coordinate systems that will ALL be used in those geometry tables (you will find even mixed types of coordinate systems!); a description of the mangle product by A. Hamilton can be found here. In the future, the CAS will probably incorporate all the necessary gemoetry information into tables and provide a user guide to utilize these tables.

Constraint on Quasar Lifetime

We have used the Jing (1998) analytical formula to compute the linear bias. In principle one can use the bias computed from numerical simulation results. However, to interpret our measured correlation function we need a fine redshift grid. Hence using the analytical bias formula is convenient. We have compared the analytical formula with that computed from numerical simulations at two redshifts: z=3 and z=4; and the analytical bias formula is a proper prescription for our integrated correlation function (see Appendix C of the paper for more discussions). Also, we compared the Press-Schechter formula and the Sheth-Tormen formula for the halo mass function with numerical simulation at z=3 and z=4. It is found that the ST formula is a better prescription for the dark matter halo mass function. The numerical simulations we used were kindly provided by Paul Bode & J. P. Ostriker, and the related figures are not included in the paper.

HMF_compare_z3.eps : halo mass function at z=3 (y axis units should be h^4 Mpc^{-3} M_{\odot}^{-1})
HMF_compare_z4.eps : halo mass function at z=4 (y axis units should be h^4 Mpc^{-3} M_{\odot}^{-1})
comp_halo_mass_CF_Paul_z3_z4.eps : comparsion of the DM CF and halo (M>2e12 h^{-1} M_{\odot}) CF with results computed using analytical CDM power spectrum and linear bias formula.