BOSS Galaxy Target Selection

Overview

This page summarizes the two principal BOSS galaxy samples, LOWZ and CMASS, as well as the smaller auxiliary samples, CMASS Sparse and CMASS Commissioning. Look here to learn how these galaxies were selected and how to select these subsamples using the target selection flag BOSS_TARGET1. For more details please refer to the Galaxy Target Selection paper: Padmanabhan et al. 2012

Whilst reading this page please keep in mind the following important, and occasionally subtle, points:

  • Unless specified differently, in what follows all colors are computed using model magnitudes, and magnitude cuts are applied on cmodel magnitudes. All quantities are corrected for Galactic extinction following Schlegel, Finkbeiner & Davis (1998).
  • A galaxy may be in more than one sample.
  • For chunks boss1 to boss14 the targeting photometry of a given object may not correspond to the final photometry of the same object. This affects a tiny percentage of targets, and may mean that the final photometry of a target falls outside the color and flux limits. In these cases, such objects should still be considered as valid targets - the scatter across the boundaries simply reflects the stochastic element of targeting a sample from noisy data. To find the original targeting photometry for any galaxy use the targetObjID in specObj (either within CAS or within the flat files).
  • Early data taken by BOSS (chunks boss1 and boss2) used a different definition of the BOSS_TARGET1 flags. In particular, the GAL_CMASS and GAL_CMASS_SPARSE bits were used for internal tests and should not be used to select objects from these chunks. In order to select a CMASS or CMASS_SPARSE sample of objects, one should select based on the GAL_CMASS_COMM bit and sub-select objects that pass the final CMASS cuts (taking into account possible changes in photometry) or, alternatively, exclude chunks boss1 and boss2 (~5% of the data) - more details below.

 

The BOSS LOWZ Galaxy Sample

BOSS targets low redshift (z <~ 0.4) galaxies using a set of color-magnitude cuts that follow the predicted evolution of a passively evolving stellar population with redshift. The selected galaxies are the brightest and reddest of the low-redshift galaxy population, and the targeting cuts are similar to those of SDSS-I/II Cut-I Luminous Red Galaxies (LRGs, Eisenstein et al. 2001). The LOWZ sample extends the SDSS-I/II LRGs by selecting fainter galaxies, thereby increasing the number density.

The selection cuts are as follows:

  1. |cperp| < 0.2
  2. r < 13.5 + cpar/0.3
  3. 16 < r < 19.6
  4. rpsf - rcmod > 0.3

where we have used the following auxiliary colors:

  • cpar = 0.7(g - r) + 1.2[(r - i) - 0.18)]
  • cperp = (r - i) - (g - r)/4.0 - 0.18

and rpsf is the PSF magnitude.

Cut (1) defines the color boundaries of the sample, around a passive stellar population; cut (2) is a sliding magnitude cut, designed to select the brightest galaxies at each redshift; cut (3) defines the faint and bright limits, and cut (4) performs a star-galaxy separation.

The LOWZ sample can be selected from the DR9 data by requiring the following:

  • BOSS_TARGET1 && 20
  • SPECPRIMARY == 1
  • ZWARNING_NOQSO == 0

IMPORTANT: Due to a bug in the target selection, LOWZ galaxies were incorrectly targeted during the initial stages of the survey (affecting all data taken until June 2010). In order to select a uniformly targeted sample of LOWZ galaxies one must therefore also require the following:

  • TILEID >= 10324

The BOSS CMASS Galaxy Sample

BOSS targets high redshift (0.4 < z < 0.7) galaxies using a set of color-magnitude cuts. The CMASS selection is similar in ethos to the algorithms used to target SDSS-I/II Cut-II (Eisenstein et al. 2001) and 2SLAQ LRGs (Cannon et al. 2006), using (g-r) and (r-i) colors to isolate high redshift galaxies. A crucial difference, however, is the extension of the galaxy selection towards the blue: the color boundaries that isolate red objects in SDSS-I/II CutII and 2SLAQ LRGs have been removed entirely. Finally, the CMASS selection algorithm introduces a sliding cut in color vs. magnitude specifically tuned to select the more massive objects in as uniform way as possible as a function of redshift.

The selection cuts are as follows:

  1. dperp > 0.55
  2. i < 19.86+ 1.6(dperp - 0.8)
  3. 17.5 < i < 19.9
  4. r - i < 2
  5. ifib2 < 21.5
  6. ipsf - imod > 0.2 + 0.2(20.0 - imod)
  7. zpsf - zmod > 9.125 - 0.46 zmod

Where we define the auxiliary color:

  • dperp = (r-i) - (g-r)/8.0

and ifib2 is the fiber magnitude.

Cut (1) isolates high-redshift objects; cut (2) is a sliding magnitude cut that selects the brightest or more massive galaxies with redshift; cut (3) defines the faint and bright limits; cut (4) protects from some outliers; cut (5) ensures a high redshift measurement success rate; and cuts (6) and (7) perform a star-galaxy separation.

The CMASS sample can be selected from DR9 data with the following:

  • BOSS_TARGET1 && 21
  • SPECPRIMARY == 1
  • ZWARNING_NOQSO == 0
  • (CHUNK != "boss1") && (CHUNK != "boss2")

IMPORTANT: The last line removes the first two BOSS chunks. This is necessary as the meaning of the BOSS_TARGET1 flags in these 2 chunks differs from the current (and final) definitions. For further details please see the spectro caveats page.

IMPORTANT: Note that for chunks up to and including boss14, the brightness cut for CMASS targets was ifib2 < 21.7. To improve the redshift measurement success rate, Cut (5) was later revised to the current value of 21.5. Therefore, to select a uniformly targeted sample of CMASS galaxies one must also require:

  • ifib2 < 21.5

For a small number of targets in chunks boss1 to boss14 the final photometry differs from the photometry used in the targeting (see note at the start of this page). As such, for around 1% of the targets the above cut will effectively be done in the wrong photometry. Avoid this issue by using the targetObjID in specObj to access the targeting photometry. Note that the LSS CMASS galaxy catalogue takes all of these issues into account, and contains a uniform sample of CMASS galaxies.

The BOSS CMASS Sparse Sample

The CMASS selection is extended into fainter and bluer galaxies (corresponding to lower stellar mass objects) via the CMASS Sparse sample. The CMASS Sparse cuts are largely the same as the CMASS selection cuts, with the exception of Cut (2) which instead reads:

  • 19.86 + 1.6(dperp - 0.8) < i < 20.14 + 1.6(dperp - 0.8)

As the name indicates, CMASS Sparse is sparsely sampled, to yield approximately 5 objects per deg2, corresponding to randomly selecting 1 in around 10 targets.

The CMASS Sparse sample can be selected from DR9 data with the following:

  • BOSS_TARGET1 && 23
  • SPECPRIMARY == 1
  • ZWARNING_NOQSO == 0
  • (CHUNK != "boss1") && (CHUNK != "boss2")

The last line removes the first two BOSS chunks. This is necessary as the meaning of the BOSS_TARGET1 flags in these 2 chunks differs from the current (and final) definitions. For further details please see the spectro caveats page

The BOSS CMASS Commissioning Sample

A different, more inclusive set of cuts was used during the commissioning stages of the survey to investigate possible CMASS targeting strategies. This sample of galaxies is frozen - i.e., no more GAL_CMASS_COMM have been observed since the commissioning period.

The CMASS commissioning selection cuts are as follows:

  1. dperp > 0.55
  2. i < 20.14+ 1.6(dperp - 0.8)
  3. 17.5 < i < 20.0
  4. r - i < 2
  5. ifib2 < 22
  6. ipsf - imod > 0.2 + 0.2(20.0 - imod)

Note that the commissioning cuts are the most inclusive set - galaxies flagged as GAL_CMASS_COMM are also flagged as GAL_CMASS and GAL_CMASS_SPARSE.