A collection of public spectroscopic redshifts is available in this data release.  The spec-z’s are collected from several surveys listed below.  Each survey has its own flagging scheme to indicate redshift confidence and we homogenized the flags in order to select objects with secure redshifts easily.  The spectroscopic objects are matched to the HSC objects by position and the positional and i-band magnitude differences of the matched objects are stored in the specz table.  They can be used to check if the matching is OK.  There are also boolean bits for each object to indicate in which survey observed the object.

If you use this table in your publication, please acknowledge the original data sources!

Some objects are observed in multiple surveys and we have merged such duplicated objects into one in the following manner:

• cross-match with HSC objects within 0.5 arcsec
• if there are multiple spec-z sources matched to the same HSC object, check their redshift flags
  • if none of them is secure, set z=-9 and zerr=-9
  • if one of them is secure, use that redshift and error
  • if there are >1 secure redshifts, take the mean redshift
• in all cases, merged object ID will look like “SDSS-DR10-xxx,VIPERS-PDR-yyy,WIGGLEZ-DR2-zzz”  so that you can identify the original objects.  The original flags are stored in the same way.

• GAMA does not have i-band mags.
• Some objects in PRIMUS have imag=-1 with secure redshift flag.
• In PRIMUS, objects classified as STARS are flagged INSECURE in our table for now because the redshift flags are not provided for those objects.
• In GAMA and WiggleZ, we arbitrarily assigned redshift errors.  Do not use them.

Each spectroscopic surveys has its own selection functions.  Please refer to the references below.  This table also summarizes how we define secure redshifts in our database table.

Survey	References	Notes
zCOSMOS DR3	Lilly et al. 2009	Flag 4’s, 3’s, 2.5, 2.4, 1.5, 9.5, 9.4, 9.3 are considered secure.
UDSz	Bradshaw et al. 2013 McLure et al. 2013	It seems only the secure redshifts are currently available from this survey.
3D-HST	Skelton et al. 2014 Momcheva et al. 2016	The best redshift catalog is a combination of spec-z + grism-z + photo-z catalog.  We have selected grism-z only by applying z_best_s=2.  We assume that the spec-z objects are covered by other spec-z samples included in this page.  3D-HST overlaps with HSC in COSMOS, AEGIS, and UDS, and data in those fields are made available in the database.  Grism redshift uncertainties are asymmetric (they are basically very high quality photo-z), but the uncertainty in the database is quoted as (z_err68_max – z_err68_min) * 0.5 for simplicity.
FMOS-COSMOS	Silverman et al. 2015	Redshifts with flag>=3 are considered secure.
VVDS	Le Fevre et al. 2013	Secure if flag=3, 4, 22, 24 (2x’s are secondary), 13, 14 (broad-line AGN), 213, or 214 (broad-line AGN, secondary).
VIPERS PDR1	Garilli et al. 2014	Secure if 2.6<flag<4.6, 22.6<flag<24.6 (secondary), 12.6<flag<14.6 (broad-line AGN), 212.6<flag<214.6 (broad-line AGN, secondary)
SDSS DR12	Alam et al. 2015	Secure if zWarning=0.
GAMA DR2	Liske et al. 2015	Secure if NQ>=4.  zs_err are arbitrarily set to 0.0005*(1+z) in our table.  Do not use them.
WiggleZ DR1	Drinkwater et al. 2010	Secure if Qop>=4. zs_err are arbitrarily set to 0.0003*(1+z) in our table.  Do not use them.
DEEP2 DR4	Davis et al. 2003 Newman et al. 2013	Secure if ZQUALITY>=3 and not a star.
PRIMUS DR1	Coil et al. 2011 Cool et al. 2013	Secure if ZQUALITY>=3 and not a star (no confidence flag is given to stars).