As part of DR4, we include a set of files describing the two-dimensional selection function of the survey, which gives the probability (as a function of position on the sky) that an object meeting the DEEP2 target-selection criteria (i.e. magnitude limit and color cut, as applicable) is selected for observation and successfully yielded a redshift. The calculation of this selection function (or “window function”) map has been described in Cooper et al. (2006), Coil et al. (2008), Newman et al. (2013), and references therein. These maps are provided for each of the DEEP2 fields; for a summary of the field locations and spectroscopic coverage, refer here.
The two-dimensional selection function maps take into account the actual placement and geometry of the overlapping slit masks used for DEEP2, as well as vignetting in the DEIMOS camera and the locations of gaps between the DEIMOS CCDs. In producing these maps, we account for the multi-pass nature of DEEP2 targeting and the fact that some overlapping DEEP2 slitmasks were designed simultaneously and some were not (as the mask design code will not place an object on multiple masks designed at the same time and in the same row of masks. The redshift completeness for red galaxies (with apparent R–I > 0.5) is used to determine the probability that a targeted object yields a secure redshift; the redshift success rate for bluer objects is not well correlated with observing statistics such as seeing, so the red galaxy completeness provides the best estimate we have that a targeted galaxy at a redshift where DEEP2 can obtain a redshift will actually yield a reliable redshift measurement.
In the selection function maps, we also mask out all regions where the photometric data are affected by either saturated stars or CCD defects and hence no galaxies were targeted. The region where these maps are non-zero provides a geometrical outline of the regions where DEEP2 targeted galaxies on the sky, while the actual value at a given position represents the probability that we both targeted a galaxy (which depends on how many masks overlap a given area, whether the object is in the first or second-pass region of those masks, and whether they were designed simultaneously) and then successfully measured its redshift. Note that the provided files do not include pointing 14 in the northernmost region of the EGS. Due to poorer photometry, the mask design differed in that pointing from the rest of the survey, such that the spatial selection is not uniform between it and the rest of DEEP2. No DEEP2 clustering measurements use data from that pointing. Improved photometry for this region is coming soon, based on imaging obtained as part of the CFHT Legacy Survey.
The selection function maps are distributed in the form of FITS images, with World Coordinate System headers describing the mapping from right ascension and declination to pixel, and the value at a given pixel (ranging from 0 to 1) being the combined selection and redshift success probability for a DEEP2 target galaxy at that position. Outside of the EGS, where the mask-making was done independently for each CFHT pointing, we provide one file per pointing; for the EGS, we provide one file describing the selection
function over the entire field. The layout of the DEEP2 “fields” and photometric “pointings” are illustrated here. In order to produce random catalogs for calculating correlation functions, one should use the value of the selection function at a given position as the probability of keeping an object (between 0 and 1) at that location within the catalog. We note that for some measurements, it may also be necessary to correct for the dependence of selection probability on source density at small scales (see Section 8 of Newman et al. 2013).
Selection Function FITS Files are available for each DEEP2 field/pointing:
accessible via Google Drive here
For more details regarding the DEEP2 target-selection procedures and the construction of the 2-d selection function maps, refer to Newman et al. (2013).