The Gemini/HST Galaxy Cluster Project: Redshift 0.2–1.0 Cluster Sample, X-Ray Data, and Optical Photometry Catalog

Galaxy evolution can be studied through observations of galaxies at different redshifts. Systematic surveys of clusters published in the mid to late 1990s investigated the evolution of the galaxy population out to z ≈ 1 using photometric measurements, in some cases combined with low-resolution spectroscopic data. Examples include the Canadian Network for Observational Cosmology (CNOC) surveys (Yee et al. 1996, 2000) and the "MORPHS" project led by Smail and Dressler (Smail et al. 1997; Dressler et al. 1999).

The goal of the CNOC cluster survey was to establish the mass distribution within the clusters. However, the data, combined with CNOC2 field galaxy data, were also used for investigations of the evolution of galaxies from z ≈ 0.6 to the present. Schade et al. (1996a, 1996b) studied the evolution of luminosities as a function of redshift and sizes, and tested for environmental effects. The results supported passive evolution for bulge-dominated galaxies, and showed no environmental dependence of the evolution of disk- or bulge-dominated galaxies. Balogh et al. (1997, 1998) focused on the star formation rates (SFRs) as measured from the [O ii] emission lines and demonstrated the significantly lower SFRs present in cluster disk galaxies compared to those in similar galaxies in the field.

The MORPHS project provided imaging with Hubble Space Telescope (HST) (Smail et al. 1997) and low-resolution spectroscopy (Dressler et al. 1999) of 10 clusters at z = 0.37–0.56. The data have been used for studies of morphological evolution (Dressler et al. 1997), evolution of (U − V) colors with redshift (Ellis et al. 1997), as well as studies of star formation history. In particular, Dressler et al. (2004) used stacked MORPHS spectra, combined with similar data for higher-redshift clusters, to establish that younger stellar populations were present in the higher-redshift clusters.

With increased access to 8 m class telescopes, a number of surveys focused on more detailed studies of the spectral properties of the cluster galaxies have been carried out. The European southern Observatory (ESO) large project ESO Distant Cluster Survey (EDisCS) targeted clusters at z = 0.4–0.9 (White et al. 2005). Based on these data, Sánchez-Blázquez et al. (2009) studied the stellar populations from stacked spectra, while Saglia et al. (2010) investigated the size evolution and established the fundamental plane (FP; Dressler et al. 1987; Djorgovski & Davis 1987) for the clusters. The results support passive evolution of bulge-dominated galaxies, but also indicate that a large fraction of the now passive galaxies entered the red sequence between z ≈ 0.8 and ≈0.4.

The Gemini Cluster Astrophysics Survey (GCLASS) consists of spectroscopic follow-up of 10 of the richest z ≈ 1.1 clusters from the Spitzer Adaptation of the Red Sequence Survey (SpARCS) survey (Muzzin et al. 2009; Wilson et al. 2009). One of the key results from GCLASS concerns the relative roles of environment or galaxy mass as the driver of the evolution of the galaxies (Muzzin et al. 2012). These authors conclude that the environment primarily affects the fraction of star-forming galaxies, while the galaxy mass determines the stellar populations.

Beyond z ≈ 1, deep spectroscopic observations become very challenging. The survey GOGREEN (Gemini Observations of Galaxies in Rich Early ENvironments) aims to study the stellar populations of both red sequence and star-forming galaxies, and to cover a large range in galaxy masses (Balogh et al. 2017). The project includes 12 clusters and nine groups at z = 0.8–1.5. The spectroscopy is of sufficient spectral resolution to study absorption lines, but cannot be used to determine the velocity dispersions of the galaxies.

Another approach is to use primarily imaging data at these redshifts. For example, the HAWK-I Cluster Survey (PI: Lidman) covers nine clusters at z = 0.8–1.5 with near-IR imaging obtained with the VLT; see the project summary in Cerulo et al. (2016). The project aims to study galaxy populations of z > 0.8 clusters, primarily from multiband photometry. The sample includes some of the most massive known clusters at these redshifts.

The above brief summary of large projects is by no means a complete list of past and ongoing effort, but serves to show examples of the different approaches taken in this field. Ultimately, all of the projects aim to establish aspects of the galaxy evolution from high redshift to the present by quantifying the galaxy properties at different redshifts.

Our project, the Gemini/HST Galaxy Cluster Project (GCP) shares this aim. The clusters in the GCP are significantly more massive than the bulk of the clusters in EDisCS and GOGREEN. The GCP data include multiband optical photometry obtained with Gemini, high-resolution imaging primarily from HST, and deep ground-based optical spectroscopy. Our spectroscopic observations have signal-to-noise ratio (S/N) higher than those reached by other projects covering similar redshifts, and have sufficient spectral resolution for reliable measurements of velocity dispersions and absorption line indices for individual galaxies. The original GCP, which is the topic of this paper, covers z = 0.2–1 (Jørgensen & Chiboucas 2013; Jørgensen et al. 2017). Our high-redshift extension of the project, xGCP, is aimed at z = 1.2–2.0. The first results for z > 1 galaxies include measurements of galaxy velocity dispersions and line strengths, and we establish for the first time the FP for a significant cluster sample at z = 1.3 (Jørgensen et al. 2014). Future papers will provide more detail and results for the xGCP.

In this paper, we present the X-ray data and catalog of the ground-based photometry for the z = 0.2–1.0 GCP clusters. We start by describing the main science goals, methods, and observing strategy of the GCP in Section 2. The section also details the cluster selection, contains an overview of previously published papers originating from the project, describes the calibration of the cluster X-ray, data and summarizes the properties of the GCP clusters. The processing of the ground-based imaging and the determination and calibration of the photometry are covered in Sections 3–4. Comparisons with photometry from the Sloan Digital Sky Survey (SDSS) are used to ensure consistently calibrated magnitudes and colors. Section 5 presents the fully calibrated photometry. The catalog is available as a machine-readable table. In our analysis of the GCP data, we make use of photometry calibrated to the rest-frame B band as well as to the rest-frame colors (U – B) and (B – V). These calibrations are established in Section 6. In Sections 7 and 8, we describe the sample selection for our spectroscopic observations, establish the color–magnitude relations in the observed bands, and finally discuss the evolution of the color–magnitude relations as a function of redshift. Section 9 summarizes the paper.

Throughout this paper, we adopt a ΛCDM cosmology with ${{\rm{H}}}_{0}=70\,\mathrm{km}\,{{\rm{s}}}^{-1}\,{\mathrm{Mpc}}^{-1}$, Ω_M = 0.3, and Ω_Λ = 0.7.

2.1. Science Goals, Cluster Sample, Observing Strategy, and Methods

The Gemini/HST Galaxy Cluster Project (GCP) was designed to study the evolution of bulge-dominated passive galaxies in very massive clusters. The main scientific goals of the project are to investigate to what extent these galaxies share a common evolutionary path and to map such a path. In the process, we can quantify the dependence on galaxy properties and possibly on cluster properties. This present paper serves as the main reference for GCP cluster selection, project description, X-ray data, and the catalog of the ground-based photometry.

The original cluster selection was based on X-ray luminosities and spectroscopic redshifts as available in the literature in the period 2000–2004. Fifteen clusters were selected for the project, with the aim to have three to four clusters for each 0.2 interval in redshift from z = 0.2 to z = 1.0. MS 1610.4+6616, selected as a cluster at z = 0.65, turned out not to be a massive cluster. The apparently extended X-ray emission detected by the Einstein satellite likely originates from several point sources. Our spectroscopic data of galaxies in this field show no well-defined concentration in redshift space consistent with a massive cluster. This leaves us with 14 clusters, and also the effect that the redshift interval z = 0.6–0.8 is rather sparsely covered by our sample.

Using the X-ray data from Piffaretti et al. (2011), the luminosity limit for the sample is ${L}_{500}={10}^{44}\,\mathrm{erg}\,{{\rm{s}}}^{-1}$ in the 0.1–2.4 keV band and within the radius R₅₀₀. The radius R₅₀₀ is the radius within which the average cluster overdensity is 500 times the critical density of the universe at the redshift of the cluster. The cluster properties are summarized in Table 1, including information on L₅₀₀, R₅₀₀, and the corresponding masses M₅₀₀.

Table 1.  Cluster Properties

Cluster	Redshift	σcluster	L500	M500	R500	Nmember	References
		km s−1	1044 erg s−1	1014 M☉	Mpc
(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Abell 1689/RX J1311.4–0120	0.1865 ± 0.0010	${2182}_{-163}^{+150}$	12.524	8.392	1.350	72	J2017
RX J0056.2+2622/Abell 115	0.1922 ± 0.0008	${1444}_{-140}^{+119}$	7.485	6.068	1.206	58	J2017
RX J0056.2+2622Na	0.1932 ± 0.0010	${1328}_{-334}^{+213}$	3.935	4.100	1.058	12	J2017
RX J0056.2+2622Sa	0.1929 ± 0.0010	${1218}_{-206}^{+164}$	4.094	4.200	1.067	22	J2017
RX J0142.0+2131	0.2794 ± 0.0009	${1283}_{-138}^{+166}$	5.587	4.761	1.079	30	B2005
RX J0027.6+2616	0.3650 ± 0.0009	${1232}_{-165}^{+122}$	8.376	5.684	1.108	34	J2017
RX J0027.6+2616 group	0.3404 ± 0.0003	${172}_{-47}^{+29}$	⋯	⋯	⋯	9	J2017
Abell 851	0.4050 ± 0.0008	${1391}_{-112}^{+102}$	4.907	3.970	0.980	50	H2018
RX J1347.5–1145b	0.4506 ± 0.0008	${1259}_{-250}^{+210}$	8.278	5.264	1.046	43	J2017
RX J2146.0+0423	0.532	⋯	1.912	2.015	0.736	⋯	In preparation
MS 0451.6–0305	0.5398 ± 0.0010	${1450}_{-159}^{+105}$	15.352	7.134	1.118	47	J2013
RX J0216.5–1747	0.578	⋯	2.267	2.147	0.738	36	In preparation
RX J1334.3+5030	0.620	⋯	3.406	2.656	0.779	⋯	In preparation
RX J1716.6+6708c	0.809	⋯	4.368	2.623	0.718	⋯	In preparation
MS 1610.4+6616 fieldd	0.8300 ± 0.0011	${681}_{-195}^{+132}$	⋯	⋯	⋯	12	In preparation
RX J0152.7–1357	0.8350 ± 0.0012	${1110}_{-174}^{+147}$	6.291	3.222	0.763	29	J2005
RX J0152.7–1357Nb	0.8372 ± 0.0014	681 ± 232	1.933	1.567	0.599	7	J2005
RX J0152.7–1357Sb	0.8349 ± 0.0020	866 ± 266	2.961	2.043	0.657	6	J2005
RX J1226.9+3332	0.8908 ± 0.0011	${1298}_{-137}^{+122}$	11.253	4.386	0.827	55	J2013
RX J1415.1+3612e	1.0269 ± 0.0010	${676}_{-77}^{+69}$	7.773	3.109	0.698	18	In preparation

Notes. Column 1: galaxy cluster. Column 2: cluster redshift. Column 3: cluster velocity dispersion. Column 4: X-ray luminosity in the 0.1–2.4 keV band within the radius R₅₀₀. X-ray data are from Piffaretti et al. (2011), except as noted. Column 5: Cluster mass derived from X-ray data within the radius R₅₀₀. Column 6: radius within which the mean overdensity of the cluster is 500 times the critical density at the cluster redshift. Column 7: number of member galaxies for which spectroscopy has been obtained. Column 8: references for redshifts, velocity dispersions, and spectroscopic data. B2005—Barr et al. (2005), updated to use a consistent method for determining the velocity dispersion; J2005—Jørgensen et al. (2005); J2013—Jørgensen & Chiboucas (2013); J2017—Jørgensen et al. (2017); H2018 P. Hibon et al. (2018, in preparation). In preparation: papers in preparation. Except for MS 1610.4+6616 and RX J1415.1+3612, our spectroscopic data for these clusters are not fully processed. Thus, we do not list the cluster velocity dispersions.

^aRecalibrated X-ray data from Mahdavi et al. (2013, 2014); see Section 2.3. ^bRecalibrated X-ray data from Ettori et al. (2004); see Section 2.3. ^cAverage of recalibrated X-ray data from Ettori et al. (2004, 2009); see Section 2.3. ^dMS 1610.4+6616 is not a rich galaxy cluster. There are 27 galaxies in the redshift interval 0.60–0.86, 12 of which are clustered at z = 0.83. ^eAverage of recalibrated X-ray data from Ettori et al. (2009), Stott et al. (2010), and Pascut & Ponman (2015); see Section 2.3.

Download table as:  ASCIITypeset image

For each cluster, we have obtained ground-based imaging in three or four of the passbands g', r', i' and z'. The photometry typically reaches a limiting magnitude of 25 mag in the passband closest to the B band in the rest frame of the clusters. The photometry is used for (1) sample selection for spectroscopic observations and (2) calibration of both the ground-based photometry and photometry from higher spatial resolution imaging to rest-frame magnitudes and colors.

The spectroscopic samples contain 30–60 candidate members in each cluster. This usually results in spectroscopic data for 20 or more passive bulge-dominated members in each cluster. The S/N and resolution of the spectra are sufficient to reliably measure velocity dispersions and absorption line strength for individual galaxies. Our samples span from the brightest cluster galaxies with typical dynamical masses of ≈ 10^12.6 M_☉ to galaxies with dynamical masses of  ≈ 10^10.3 M_☉, equivalent to a velocity dispersion of about 100 km s⁻¹. All collection of ground-based imaging and spectroscopy was done in the period 2001–2005 with Gemini North and South, using the Gemini Multi-Object Spectrographs GMOS-N and GMOS-S. See Hook et al. (2004) for a description of the instruments.

The GCP makes use of high spatial resolution imaging of the clusters primarily from the Advanced Camera for Surveys (ACS) or the Wide Field and Planetary Camera 2 (WFPC2) on board HST. The data are in part archive data obtained for other programs and in part a result of our approved programs. RX J0056.2+2622 was covered by high-resolution ground-based imaging in the r' band obtained with Gemini North. High spatial resolution imaging is used to measure half-light radii, mean surface brightnesses, and total magnitudes from fits with Sérsic profiles (Sérsic 1968) and r^1/4 profiles. The Sérsic indices are used to ensure that our final samples for the analysis are indeed bulge-dominated galaxies. The details of our methods for determining these parameters can be found in Chiboucas et al. (2009). Table 2 gives an overview of the relevant HST data. Additional two-dimensional photometry derived from these data will be included in future papers. For some of the clusters, data are available for filters with wavelengths shorter than those listed in the table, but these are not used in the GCP.

Table 2.  Relevant Hubble Space Telescope Imaging data

Cluster	Program ID	Instrument	Filters	Exptime	References
(1)	(2)	(3)	(4)	(5)	(6)
Abell 1689	9289	ACS	F625W, F775W, F850LP	9500, 11800, 14200	J2018
Abell 1689	11710	ACS	F814W	75180	J2018
Abell 1689	5993	WFPC2	F602W, F814W	1800, 2300	J2018
RX J0142.0+2131	9770	ACS	F775W	4420	C2009
RX J0027.6+2616	9770	ACS	F775W	7185	J2018
Abell 851	10418	ACS	F814W	5464	H2018
Abell 851	5190, 6480	WFPC2	F702W	4400–18900	H2018
Abell 851	5378	WFPC2	F814W	12600	H2018
RX J1347.5–1145	12104	ACS	F606W, F625W, F775W	1939, 1924, 2048	J2018
RX J1347.5–1145	10492, 11591	ACS	F814W, F850LP	7340, 5280	J2018
RX J2146.0+0423	10152	ACS	F814W	2150	In preparation
MS 0451.6–0305	9836	ACS	F814W	4072	JC2013
RX J0216.5–1747	9770	ACS	F775W	10120	In preparation
RX J1334.3+5030	9770	ACS	F775W	10120	In preparation
RX J1716.6+6708	7293	WFPC2	F555W, F814W	5600, 2700	In preparation
MS 1610.4+6616	10826	WFPC2	F702W	24000	In preparation
RX J0152.7–1357	9290	ACS	F625W, F775W, F850LP	4750, 4800, 4750	C2009
RX J1226.9+3332	9033	ACS	F606W, F814W	4000, 4000	C2009
RX J1415.1+3612	10496	ACS	F850LP	9920	In preparation

Note. Column 1: cluster name. Column 2: HST program ID. Column 3: HST instrument. Column 4: filters for imaging. Column 5: total exposure times in seconds for each of the filters. Column 6: reference for our two-dimensional photometry. C2009—Chiboucas et al. (2009), JC2013—Jørgensen & Chiboucas (2013), J2018—I. Jørgensen et al. (2018, in preparation), H2018—P. Hibon et al. (2018, in preparation).

Download table as:  ASCIITypeset image

Our main methods for analysis so far have been to (1) study how the scaling relations like the FP and velocity dispersion—line strength relations evolve with redshift, and (2) investigate the distributions of ages, metallicities, and abundance ratios as well as establish how these parameters depend on galaxy velocity dispersion, redshift, and possibly the cluster environment. Our previous papers detail the results of this analysis of eight of the clusters, see Jørgensen et al. (2005, 2006, 2007, 2017), Barr et al. (2005), Jørgensen & Chiboucas (2013). The next section provides a brief overview of these and other papers relevant for the project.

2.2. Previous Papers from the GCP

2.3. Calibration of X-Ray Data

The comprehensive X-ray cluster catalog by Piffaretti et al. (2011) provides consistently calibrated X-ray data for the majority of the GCP clusters. However, RX J1716.6+6708 and RX J1415.1+3612 are not included in this catalog, and it treats the binary clusters RX J0056.2+2622 and RX J0152.7–1357 as single clusters. To cover these clusters, we calibrate X-ray data from Ettori et al. (2004, 2009), Stott et al. (2010), Mahdavi et al. (2013, 2014), and Pascut & Ponman (2015) to be consistent with the Piffaretti et al data. In addition, we use updated (and calibrated) values for RX J1347.5–1145 from Ettori et al. (2004), who correct the X-ray measurements for diffuse emission from an infalling subcluster to the southeast of the main cluster; see also Jørgensen et al. (2017) for a discussion of this cluster.

In the calibration, we use conversions between radii R₅₀₀, masses M₅₀₀, and luminosities L₅₀₀ as given by Piffaretti et al. in their Equations (2) and (3). We reproduce them here for clarity:

$$\begin{eqnarray}&&h{(z)}^{-7/3}\left(\displaystyle \frac{{L}_{500}}{{10}^{44}\,\mathrm{erg}\,{{\rm{s}}}^{-1}}\right)=C{\left(\displaystyle \frac{{M}_{500}}{3\cdot {10}^{14}{M}_{\odot }}\right)}^{\alpha },\end{eqnarray} \tag{ 1 }$$

where h(z) is the Hubble factor at redshift z, log C = 0.274, and α = 1.64, and

$$\begin{eqnarray}&&{M}_{500}=\displaystyle \frac{4\pi }{3}\,{R}_{500}^{3}500{\rho }_{c}(z),\end{eqnarray} \tag{ 2 }$$

where ${\rho }_{c}{(z)=3H(z)}^{2}/(8\pi G)$ is the critical density of the universe at redshift z.

We convert the X-ray data from the literature to M₅₀₀. As needed, we also adopt the following conversions from Piffaretti et al.: L₅₀₀ = 0.91 L_total, R₂₀₀ = 1.52 R₅₀₀, and L₅₀₀ = 0.96 L₂₀₀. The relation between R₂₀₀ and R₅₀₀ is equivalent to M₂₀₀ = 1.40 M₅₀₀; cf. Equation (2). When other conversions are used in the literature, we remove those and apply the above conversions before comparing with data from Piffaretti et al.

We determine the offsets in log M₅₀₀ between the other catalogs and Piffaretti et al. to establish the best offset for each set of data. Table 3 and Figure 1 summarize the comparisons and the adopted offsets.

Zoom In Zoom Out Reset image size

Table 3.  Calibration of Cluster X-Ray Measurements

Catalog	Primary measure	N	Mean	Median	rms	Δ
(1)	(2)	(3)	(4)	(5)	(6)	(7)
Ettori et al. (2004)	R500, M500	14	0.23	0.23	0.21	0.23
Ettori et al. (2009)	R500, M500	31	0.24	0.29	0.23	0.29
Stott et al. (2010)	TX, M500	7	0.18	0.27	0.19	0.18
Mahdavi et al. (2013, 2014)	MHydro	29	0.02	0.00	0.14	0.00
Pascut & Ponman (2015)	R500, M500	20	0.08	0.08	0.13	0.08

Note. Column 1: reference for X-ray data. Column 2: primary parameter from catalog; see the text. Column 3: number of clusters in common with Piffaretti et al. (2011). Column 4: mean of the differences in $\mathrm{log}{M}_{500}$; differences are calculated as Catalog – Piffaretti. Column 5: median of the differences. Column 6: rms of the differences. Column 7: adopted offset in $\mathrm{log}{M}_{500}$ to reach consistency with Piffaretti et al.

Download table as:  ASCIITypeset image

2.4. Cluster Properties

Table 1 summarizes the properties for all GCP clusters. In Figure 2 we show the cluster masses versus redshifts for these clusters. For reference, the figure also shows our local reference sample and the clusters from Piffaretti et al. (2011). We show sample models for the growth of cluster masses with time, based on simulations from van den Bosch (2002). These models are in general agreement with newer and more detailed analyses of the results from the Millennium simulations (Fakhouri et al. 2010).

Zoom In Zoom Out Reset image size

Grayscale optical images with X-ray data overlaid of clusters for which we previously have published results are available in those papers as follows: RX J0152.7–1357 is published in Jørgensen et al. (2005), MS 0451.6–0305 and RX J1226.2+3332 are published Jørgensen & Chiboucas (2013), and Abell 1689, RX J0056.2+2622, RX J0027.6+2616, and RX J1347.5–1147 are published in Jørgensen et al. (2017). The remaining clusters are shown in Appendix C of the current paper, Figures 20–27. The grayscale images show the spectroscopic samples, and when available at this time, information about cluster membership and galaxy properties. In Appendix C, we also describe each of the clusters, including providing the original references for their discovery and main references for substantial results on the cluster properties and, when available, the star formation history of their members.

Ground-based imaging of the clusters was obtained with GMOS-N and GMOS-S in the period from 2001 to 2005. Each cluster was observed in three or four of the filters g', r', i' and z'. Table 4 summarizes the instrument information, while Table 16 in Appendix A gives detailed information on the available observations, including the Gemini program IDs. That table also lists the adopted Galactic extinction for each of the fields and filters. Abell 1689 and RX J0056.2+2622 were observed with two pointings, while all other clusters have data for one pointing.

Table 4.  Instrumentation

Parameter	GMOS-N	GMOS-S
CCDs	3 × E2V 2048 × 4608	3 × E2V 2048 × 4608
r.o.n.a	(3.5, 3.3, 3.0) e−	(4.0, 3.7, 3.3) e−
gaina	(2.04, 2.3, 2.19) e−/ADU	(2.33, 2.07, 2.07) e−/ADU
Unbinned pixel scale	0.0727arcsec/pixel	0.073arcsec/pixel
Field of view	55 × 55	55 × 55
Imaging filters	$g^{\prime} ,r^{\prime} ,i^{\prime} ,z^{\prime}$	$g^{\prime} ,r^{\prime} ,i^{\prime} ,z^{\prime}$

Note.

^aValues for the three detectors in the array.

Download table as:  ASCIITypeset image

3.1. Processing of Imaging Data

3.2. Derived Photometric Parameters

The co-added images were processed with SExtractor version 2.8.6 (Bertin & Arnouts 1996). We used SExtractor in dual-image mode, with the images pre-registered to each other. The image in the filter closest to the rest-frame B band was used for detections, while the images in the other filters were used only for photometry. For consistency between clusters, the threshold for detection was defined as a surface brightness. The detection thresholds combined with the requirement of meeting the threshold over a minimum of 9 pixels correspond to an S/N of 8–10. The analysis threshold in the other bands was then defined using the approximate expected colors of the cluster members on the red sequence. In all cases, we maintain analysis thresholds corresponding to an S/N of 5–6 or better over the minimum detection area of 9 pixels. Thus, for objects on the red sequence, roughly the same aperture size in each band is used to derive the geometrical parameters. The geometrical parameters are used by SExtractor to derive the class_star parameter, which we use to separate galaxies and stars. Table 5 lists the filter used for detection and the adopted thresholds.

Table 5.  Photometry Overview

Cluster	Detection band	Thresholds	Apertures	5σ limit	Nstar	Ngalaxy	Area
(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Abell 1689	g'	25.5, 24.5, 24.0, ...	4.35	24.8	90	1981	44.9a
RX J0056.2+2622	g'	25.5, 24.5, 24.0, ...	4.18	25.1	113	991	53.3a
RX J0142.0+2131	r'	27.0, 25.5, 24.9, ...	3.08	24.8	55	1546	28.6
RX J0027.6+2616	r'	27.0, 25.5, 24.9, ...	2.66	25.2	75	1291	28.0a
Abell 851	r'	27.0, 25.5, 24.9, ...	2.61	25.2	60	1684	28.5
RX J1347.5–1145	r'	27.0, 25.5, 24.8, ...	3.09, 2.51	25.0	100	952	27.1a
RX J2146.0+0423	r'	27.0, 25.5, 24.5, ...	2.17	25.0	156	1223	28.8
MS 0451.6–0305	r'	27.0, 25.5, 24.5, 24.1	2.44	25.4	95	1453	27.5a
RX J0216.5–1747	i'	27.0, 25.5, 24.5, 24.1	2.24	24.8	49	710	26.8a
RX J1334.3+5030	i'	..., 26.3, 25.3, 24.9	2.74, 2.04	25.1	50	967	27.1a
RX J1716.6+6708	i'	..., 26.5, 25.3, 24.6	2.14	24.9	104	1126	27.2a
MS 1610.4+6616 field	i'	..., 26.3, 25.3, 24.9	2.25	25.1	98	1180	27.8
RX J0152.7–1357	i'	..., 26.5, 25.3, 24.6	2.06	24.7	51	1732	28.6
RX J1226.9+3332	i'	..., 26.5, 25.3, 24.6	2.17	25.0	61	1056	27.8
RX J1415.1+3612	z'	..., 26.5, 25.4, 24.7	2.04	24.2	76	1132	26.8b

Notes. Column 1: galaxy cluster. Column 2: filter used for detections. Column 3: thresholds in mag arcsec² in the order (g', r', i', z'). Column 4: diameter of apertures in arcsec. For RX J1347.5–1145, 3.09 arcsec was used for g' and 2.52 arcsec for r' and i', while for RX J1334.3+5030, 2.74 arcsec was used for r' and i' and 2.04 arcsec was used for z'; see the text. Column 5: 5σ detection limit in magnitudes in the detection band; see Section 3.3. Column 6: number of stars in catalog. Column 7: number of galaxies in catalog. Column 8: area observed in arcmin². Abell 1689 and RX J0056.2+2622 were both observed with two slightly overlapping GMOS-N fields. The other clusters were covered with one field. Small variations in the final area are due to differences in dither patterns and vignetting from the OIWFS as noted.

^aAreas affected by vignetting by the OIWFS are excluded from the total area. ^bAreas around bright foreground galaxies are excluded from the total area.

Download table as:  ASCIITypeset image

The SExtractor background mesh size was adjusted to avoid systematic effects from the galaxies with the largest angular size. We typically use a background mesh size of 256 pixels, with a filter size of 5 pixels. We used 64 subthresholds and a minimum contrast for the deblending of only 0.0005 (the default is 0.005). This enables deblending of objects in these fairly crowded fields. For Abell 1689 and RX J1334.3+5030, even lower minimum contrast for a deblending of 0.00002 was needed to deblend fainter objects in the vicinity of either the brightest galaxies in the cluster center (Abell 1689) or at close angular distance to bright stars (RX J1334.3+5030). For Abell 1689, the lower deblending contrast is used only within 30 arcsec of the cluster core (for this purpose defined as the position of the galaxy with ID 626). For RX J1334.3+5030, the detections were done in the i' band, which has an image quality of FWHM = 0.87 arcsec (measured as the FWHM from a Gaussian fit to stars in the image). We used the better seeing z'-band image (FWHM = 0.54 arcsec) to check that the deblending was correct. In all cases, we use the SExtractor convolution file Gauss_2.0_3x3.conv. We visually inspected all fields to ensure that galaxies in our spectroscopic samples were correctly deblended.

SExtractor was run without a weight image. However, the catalogs were cleaned of spurious detections along the edges of the field and along the edges of any vignetting from the GMOS on-instrument wavefront sensor (OIWFS) when this is inside the field of view. Table 5 lists the effective area for object detection after such cleaning.

We adopt mag_auto from SExtractor as the total magnitudes of the objects, as these magnitudes are consistently derived based on apertures 2.5 times the Kron radii, r_Kron (Kron 1980). See Graham & Driver (2005) for a discussion of the implementation of the Kron radius in SExtractor. In some cases of close neighbors, the magnitudes may be affected by these. We also provide the isophotal magnitudes, mag_iso, in the photometry table. In Section 3.4, we discuss to what extent mag_auto is different from true total magnitudes, and we derive aperture corrections for point sources.

Differences in image quality between the observations in the different passbands can complicate the determination of colors of the galaxies. Various techniques to address this issue have been used in the past. One approach is to used fixed size apertures, but to convolve all images of a given field to a common (worst) resolution, or, less drastically, convolving the images only in pairs as described by Meyers et al. (2012) and used by, e.g., Cerulo et al. (2016). Alternatively, one may obtain "global" colors of the galaxies, using mag_auto as the basis for the colors. We note that mag_auto is also the only choice of the SExtractor "total" magnitudes that use the same size aperture for all frames.

Instead of convolving the images, we take the approach of measuring the aperture colors, using aperture sizes chosen to minimize the effect of image quality differences on the measured colors. To decide on the aperture sizes, we first estimate aperture diameters in arcsec using the image quality, FWHM, and the half-light radius of typical small galaxies in the clusters. Specifically, the aperture diameter in arcsec is chosen to be

$$\begin{eqnarray}&&{D}_{\mathrm{app}}=2\cdot 2.355{({(\mathrm{FWHM}/2.355)}^{2}+{r}_{\mathrm{galaxy}}^{2})}^{0.5},\end{eqnarray} \tag{ 3 }$$

where r_galaxy is the half-light radius in arcsec at the redshift of the cluster, corresponding to a physical size of 2.5 kpc. For those clusters with differences between D_app in the different passbands of less than 10%, we then used the largest of those diameters as the aperture size for all of the passbands. For RX J1347.5–1145 and RX J1334.3+5030, the differences between D_app for the different passbands were larger than 10%. Thus, for these clusters, we used two different aperture sizes. Aperture sizes are listed in Table 5. In the catalog table (see Section 5), we give the aperture magnitudes. For reference, we also provide aperture magnitudes within an aperture with a diameter of 2.5 arcsec.

Other adjustments to the SExtractor parameters were trivial adjustments to the magnitude zero-points and the image quality.

Stars and galaxies were separated based on the SExtractor classification parameters class_star for all available bands. We derive the product and the median of those available, and define

$$\begin{eqnarray}&&P({\mathtt{class}}\_{\mathtt{star}})\equiv \prod {\mathtt{class}}\_{\mathtt{star}},\end{eqnarray} \tag{ 4 }$$

$$\begin{eqnarray}&&M({\mathtt{class}}\_{\mathtt{star}})\equiv \mathrm{median}({\mathtt{class}}\_{\mathtt{star}}).\end{eqnarray} \tag{ 5 }$$

Objects are classified as stars if they meet the criterion

$$\begin{eqnarray}&&P({\mathtt{class}}\_{\mathtt{star}})\geqslant {0.8}^{N}\,| | \,M({\mathtt{class}}\_{\mathtt{star}})\geqslant 0.8,\end{eqnarray} \tag{ 6 }$$

while all other objects were considered galaxies. N is the number of bands available for a given field. The classification of saturated stars was set manually. Figure 3 shows M(class_star) versus P(class_star) for two of the clusters, RX J2146.0+0423 with photometry in three bands and RX J0216.5–1747 with photometry in four bands. The image quality for the observations of these two clusters is comparable, 0.50–0.65 arcsec. The parameters P(class_star) and M(class_star) are included in the table of the photometric data (see Section 5), allowing users to reclassify the objects based on the available data. Figure 4 shows P(class_star) and class_star in the detection band versus magnitudes, illustrating how the use of P(class_star) aids in the classification of especially faint objects within ≈2 mag of the detection limit. In our original sample selection of targets for spectroscopic observations, we required class_star < 0.80 in the detection band. In a few cases, our refined classification would have excluded a spectroscopic sample target from the sample. In all such cases, except the Seyfert galaxy RX J1415.1+3612 ID 983, the spectra confirm that the objects are indeed stars. In Table 5, we list the number of objects classified as galaxies and as stars in each cluster field.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

While our ground-based imaging in general is not of sufficient spatial resolution to warrant detailed two-dimensional photometry, we do provide measures of sizes as an isophotal radius, r_iso, as well as position angles and ellipticities, which may be useful for sample selections for other follow-up studies of the clusters. We use the isophotal area iso_area_image determined by SExtractor in pixels to derive a circularized isophotal radius in arcseconds as

$$\begin{eqnarray}&&{r}_{\mathrm{iso}}={({\mathtt{iso}}\_{\mathtt{area}}\_{\mathtt{image}}/\pi )}^{0.5}\mathrm{pixelscale},\end{eqnarray} \tag{ 7 }$$

where "pixelscale" is the pixel scale for the image in arcsec pixel⁻¹. The surface brightnesses used for the determinations are listed in Table 5.

3.3. Uncertainties on Magnitudes

The uncertainties in the magnitudes estimated by SExtractor are based on the sky noise per pixel, σ_sky, combined with the Poisson noise of the signal from the objects. It is assumed that the noise from the sky scales with the area, A, of the aperture in pixels such that the total uncertainty on the flux, F, measured from an object can be expressed as

$$\begin{eqnarray}&&{\sigma }_{\mathrm{object}}={(A{\sigma }_{\mathrm{sky}}^{2}+F{\mathrm{gain}}^{-1})}^{1/2},\end{eqnarray} \tag{ 8 }$$

where F is in counts and "gain" is the gain of the image in e⁻/counts. Several studies have shown that these uncertainties in general are underestimated. In particular, Labbé et al. (2003) find that even for HST imaging and small apertures the effect can be a factor of 2. Due to imperfect corrections for scattered light and/or fringing, ground-based imaging often has stronger large-scale variations of the background than those typically found in HST imaging. We adopted a combination of the method used by Labbé et al. and that by Guo et al. (2013) to determine the correction factor for the noise estimates, given the sizes of the apertures.

For each field and filter combination, we mask out pixels containing signal from objects. We then place empty apertures with areas from 16 to 400 pixels, equivalent to aperture diameters of 0.65–3.3 arcsec. For the largest size apertures, we typically place 200–300 empty apertures across the masked images, while for the smaller sizes, 600–800 empty apertures were used. The background was subtracted using the sky image produced by SExtractor. We then measure the flux in each of the empty apertures. For each size aperture, we determine the scatter of the fluxes from a Gaussian fit to their distribution. As also found by Labbé et al., the scatter for a given field and filter can be parameterized as

$$\begin{eqnarray}&&{\sigma }_{i}(A)={A}^{1/2}{\sigma }_{\mathrm{sky}}({a}_{i}+{b}_{i}{A}^{1/2}).\end{eqnarray} \tag{ 9 }$$

We determine the coefficients a_i and b_i from least-squares fits. Table 6 summarizes the determinations, the sky noise per pixel σ_sky normalized to 1 s, and the resulting correction factor a_i + b_i A^1/2 for an aperture with a diameter of 2.5 arcsec. The correction factors are between 1.8 and 6.7, with a median value of 3.2. The median of the coefficients a_i is 1.25, reflecting the typical correlation of noise between pixels due to the stacking of multiple frames, cf., Labbé et al. The coefficients b_i, reflecting the typical large-scale variations in the background, have a median value of 0.13.

Table 6.  Noise Parameters

Cluster	g' Band				r' Band				i' Band				z' Band
	a	b	σsky	Factor	a	b	σsky	Factor	a	b	σsky	Factor	a	b	σsky	Factor
(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)	(14)	(15)	(16)	(17)
A1689 F1	1.42	0.196	0.048	4.41	0.85	0.280	0.071	5.13	0.78	0.163	0.210	3.26	⋯	⋯	⋯	⋯
A1689 F2	1.25	0.223	0.047	4.65	0.62	0.259	0.091	4.56	0.87	0.111	0.213	2.56	⋯	⋯	⋯	⋯
RX J0056p2p2622 F1	0.96	0.075	0.084	2.09	1.18	0.148	0.100	3.44	1.61	0.057	0.136	2.48	⋯	⋯	⋯	⋯
RX J0056p2p2622 F2	1.15	0.045	0.080	1.83	1.21	0.140	0.095	3.34	1.39	0.162	0.115	3.85	⋯	⋯	⋯	⋯
RX J0142p0p2131	1.23	0.162	0.028	3.70	0.86	0.169	0.087	3.44	1.37	0.081	0.083	2.61	⋯	⋯	⋯	⋯
RX J0027p6p2616	1.24	0.168	0.029	3.80	0.79	0.121	0.068	2.64	1.12	0.341	0.074	6.32	⋯	⋯	⋯	⋯
A0851	1.16	0.146	0.026	3.39	1.03	0.094	0.085	2.47	1.57	0.088	0.077	2.90	⋯	⋯	⋯	⋯
RX J1347p5m1145	1.24	0.249	0.038	5.04	0.97	0.130	0.117	2.96	1.27	0.193	0.098	4.21	⋯	⋯	⋯	⋯
RX J2146p0p0423	1.38	0.166	0.020	3.91	0.70	0.190	0.070	3.59	1.33	0.099	0.063	2.84	⋯	⋯	⋯	⋯
MS 0451p6m0305	1.30	0.261	0.017	5.28	0.84	0.155	0.047	3.20	1.50	0.213	0.061	4.75	1.52	0.131	0.033	3.52
RX J0216p5m1747	1.78	0.087	0.030	3.11	1.65	0.041	0.065	2.28	1.26	0.048	0.115	2.00	1.70	0.042	0.059	2.34
RX J1334p3p5030	⋯	⋯	⋯	⋯	1.28	0.201	0.030	4.33	1.14	0.063	0.072	2.09	1.50	0.053	0.045	2.30
RX J1716p6p6708	⋯	⋯	⋯	⋯	0.94	0.259	0.044	4.89	0.97	0.097	0.094	2.44	1.47	0.089	0.068	2.82
MS 1610p4p6616	⋯	⋯	⋯	⋯	1.88	0.064	0.041	2.86	1.16	0.046	0.101	1.87	1.97	0.032	0.055	2.45
RX J0152p7m1357	⋯	⋯	⋯	⋯	1.71	0.081	0.035	2.94	2.12	0.090	0.065	3.49	1.61	0.071	0.026	2.69
RX J1226p9p3332	⋯	⋯	⋯	⋯	1.69	0.094	0.026	3.12	0.78	0.198	0.054	3.80	1.44	0.066	0.126	2.44
RX J1415p1p3612	⋯	⋯	⋯	⋯	1.09	0.282	0.033	5.40	1.08	0.367	0.056	6.67	1.51	0.077	0.051	2.69

Note. Column 1: galaxy cluster and field. Column 2: g'-band noise correction coefficient a. Typical uncertainties are 0.11. Column 3: g'-band noise correction coefficient b. Typical uncertainties are 0.009. Column 4: g'-band sky noise per pixel, σ_sky, normalized to 1 s. Column 5: g'-band noise correction factor ${a}_{i}+{b}_{i}{A}^{1/2}$ for an aperture with a diameter of 2.5 arcsec. Columns 6–9: same information for the r' band. Columns 10–13: same information for the i' band. Columns 14–17: same information for the z'-band.

Download table as:  ASCIITypeset image

The uncertainties of all magnitudes were then derived using the coefficients and the relevant sizes of the apertures. Figure 5 shows uncertainties on mag_auto as a function of mag_auto for three of the clusters spanning the redshift range of the sample and illustrating the typical depth of the data as a function of redshift and passband. In Figure 6, we show the magnitude distribution of the objects for each field in the detection band. The 5σ detection limits are listed in Table 5 and marked on the figure.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

3.4. SExtractor Magnitudes versus Total Magnitudes

The SExtractor magnitudes mag_auto are known to miss a small fraction of the total flux from objects. Bertin & Arnouts (1996) determined from simulations the loss to be 0.03–0.06 magnitudes. Theoretical work by Graham & Driver (2005) shows that the fraction lost for galaxies depends on their luminosity profile. For galaxies with Sérsic (1968) profiles, the fraction lost is ≈4% for an exponential profile, increasing to ≈10% for an r^1/4 profile. However, Graham & Driver also point out that if the Kron radius (used as the basis for mag_auto) is derived from integration over only 1–2 effective radii of the galaxies, the lost flux can be substantially larger.

We first derived aperture corrections for the point sources in a standard fashion, using magnitudes derived through large apertures for bright isolated stars. The resulting aperture corrections, Δm_aper, for the adaptive aperture sizes defined in Equation (3) as well as for the fixed aperture size of 2.5 arcsec diameter are listed in Table 7. As expected, the aperture correction for the fixed aperture size is strongly correlated with the image quality, increasing from ≈0.05 mag for the best seeing images to ≈0.15 mag at a seeing of 0.8 arcsec. The flux missed from mag_auto can then be derived as

$$\begin{eqnarray}&&{\rm{\Delta }}m={\mathtt{mag}}\_{\mathtt{auto}}-({\mathtt{mag}}\_{\mathtt{aper}}+{\rm{\Delta }}{m}_{\mathrm{aper}}).\end{eqnarray} \tag{ 10 }$$

Figure 7 shows Δm versus mag_auto for the unsaturated stars in the detection bands. The median missed flux is 0.06 mag in g', i', and z'. The missing flux in the r' band is slightly higher at 0.075 mag, presumably due to differences in the point-spread functions.

Zoom In Zoom Out Reset image size

Table 7.  Aperture Corrections for Point Sources

Cluster	g' Band		r' Band		i' Band		z' Band
	Δmaper	Δm2.5	Δmaper	Δm2.5	Δmaper	Δm2.5	Δmaper	Δm2.5
(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
Abell 1689 F1	−0.024	−0.060	−0.039	−0.166	−0.044	−0.163	⋯	⋯
Abell 1689 F2	−0.028	−0.060	−0.042	−0.153	−0.031	−0.106	⋯	⋯
RX J0056.2+2622 F1	−0.043	−0.131	−0.040	−0.111	−0.036	−0.096	⋯	⋯
RX J0056.2+2622 F2	−0.061	−0.207	−0.050	−0.135	−0.101	−0.296	⋯	⋯
RX J0142.0+2131	−0.055	−0.064	−0.041	−0.052	−0.038	−0.051	⋯	⋯
RX J0027.6+2616	−0.091	−0.099	−0.074	−0.079	−0.051	−0.052	⋯	⋯
Abell 851	−0.088	−0.097	−0.064	−0.067	−0.064	−0.071	⋯	⋯
RX J1347.5–1145	−0.178	−0.280	−0.117	−0.121	−0.123	−0.124	⋯	⋯
RX J2146.0+0423	−0.094	−0.076	−0.075	−0.057	−0.071	−0.054	⋯	⋯
MS 0451.6–0305	−0.142	−0.135	−0.067	−0.064	−0.138	−0.131	−0.134	−0.129
RX J0216.5–1747	−0.114	−0.096	−0.094	−0.079	−0.076	−0.062	−0.085	−0.075
RX J1334.3+5030	⋯	⋯	−0.215	−0.258	−0.133	−0.174	−0.098	−0.079
RX J1716.6+6708	⋯	⋯	−0.158	−0.116	−0.173	−0.130	−0.174	−0.132
MS 1610.4+6616	⋯	⋯	−0.112	−0.094	−0.113	−0.094	−0.139	−0.118
RX J0152.7–1357	⋯	⋯	−0.093	−0.071	−0.112	−0.089	−0.108	−0.090
RX J1226.9+3332	⋯	⋯	−0.148	−0.116	−0.179	−0.132	−0.131	−0.104
RX J1415.1+3612	⋯	⋯	−0.107	−0.080	−0.200	−0.139	−0.120	−0.089

Note. Column 1: galaxy cluster and field. Column 2: g'-band aperture correction for adaptive aperture size. Typical uncertainties are 0.015 mag. Column 3: g'-band aperture correction for an aperture diameter of 2.5 arcsec. Typical uncertainties are 0.015 mag. Columns 4–5: same information for the r' band. Columns 6–7: same information for the i' band. Columns 8–9: same information for the z'-band.

Download table as:  ASCIITypeset image

To assess the fraction of flux lost from mag_auto of the galaxies in the observed fields, we performed detailed simulations matching 5 of the 15 clusters, spanning the relevant parameter space in redshifts, image quality, and filters. The simulations were created using Python software by R. Peterson et al. (2018, in preparation), produced during Peterson's internship with the GCP. The simulation software calls the galaxy-fitting program GALFIT (Peng et al. 2002) to make the model galaxies.

For each cluster, we made simulated images matching the three (or four) available filters. Each simulated image contains 500–1000 model galaxies. The model galaxies have distributions in total magnitudes matching the mag_auto distributions of the real data, and two-dimensional distributions in effective radii and total magnitudes, which, once convolved with the point-spread function (PSF), match the two-dimensional distributions in (mag_auto, flux_radius) of the real data. The galaxies were assumed to have Sérsic profiles. For each galaxy, values of the Sérsic indices, n_ser; ellipticities; and position angles were chosen randomly from uniform distributions. We assumed n_ser between 0.5 and 5, ellipticities between zero and 0.7, and position angles between −90 and +90 degrees. We then used GALFIT to create noiseless model images of each galaxy. Each model galaxy was convolved with the empirical PSF of the real data. The PSFs were established from 10 to 15 isolated stars in the fields. The model galaxies were randomly placed into empty images of the same size as our GMOS images and with a background level matching the real data. Thus, these images have crowding of the objects similar to the real observations, except for the very center of the clusters. Finally, noise was added, taking into account read-out noise and Poisson noise. We did not attempt to model the correlated noise due to the image stacking, or the contribution from the non-flat sky background in the real data. We do not expect these effects to contribute significantly to the fraction of lost flux.

For each cluster, we used the same seed to generate the random samples for each of the three (or four) filters. Thus, a given model galaxy will have identical n_ser, ellipticity, and position angle in the three (or four) filters. The color of the model galaxy will represent the average color of galaxies in the field at the given magnitude. SExtractor was run in dual-image mode on the simulations, with all parameters set identically to those used for the real data. The simulations use the adaptive aperture sizes, Equation (3), for aperture magnitudes and colors.

Figures 8 and 9 summarize the results from the simulations matching RX J0142.0+2131 (z = 0.28) and RX J1226.9+3332 (z = 0.89), serving as representatives for the relevant parameter space. Panels (a)–(c) of the figures show how the simulated data match the real data in magnitudes, sizes, and colors. In particular, panels (b) show the ratio $R=2{r}_{\mathrm{iso}}{r}_{\mathrm{flux}}^{-1}$ between the aperture size within which the Kron radius is determined by SExtractor (2 r_iso) and the effective radius here approximated with r_flux from SExtactor.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Panels (d)–(f) in the figures show the lost flux, Δm, as a difference between the SExtractor mag_auto and the input total magnitude. Fainter galaxies have larger Δm. However, the main drivers for the difference are the ratio R, which depends on the magnitudes of the galaxies (panels (b)), and the assumed Sérsic index. In panels (e), we show for galaxies brighter than 23 mag Δm as a function of n_ser. The points are color-coded for R larger (orange) or smaller (blue) than 3. The simulations follow the expected dependency established by Graham & Driver, and shown on the figure. To further illustrate the dependency on both R and n_ser, panels (f) show the effect as Δm versus the ratio R, color-coded for n_ser. Based on our simulations for five clusters, we conclude that there are no significant differences in Δm due to differences in filters, image quality, or redshift of the clusters. In summary, the median Δm for galaxies brighter than 23 mag and with R ≥ 3 is 0.06, 0.13, and 0.21 for n_ser ≤ 2, 2–3.5, and ≥3.5, respectively. The galaxies included in our spectroscopic samples and our investigation of the red sequence (Section 7) typically have R ≥ 3. In practice, we do not know n_ser from our ground-based data. However, we expect that galaxies on the red sequence have n_ser ≥ 2, and therefore will have Δm ≈ 0.1–0.2 mag.

In panels (g)–(h), we show the simulation results for the main color for the two clusters. In both cases, colors based on mag_auto reproduce the input colors better than colors based on aperture magnitudes, even when aperture sizes are chosen to match the seeing; cf. Equation (3). However, when evaluating which to use to investigate the colors of the real galaxies, it should also be kept in mind that the aperture magnitudes usually have lower uncertainties due to background noise. The simulations assumed no internal color gradients in the galaxies. Color gradients in the real galaxies will of course cause differences between aperture colors and colors using mag_auto. In our discussion of the color–magnitude relation for the clusters, Section 7, we show both colors.

4.1. Initial Calibration

4.2. Comparison with SDSS Photometry

We compared our photometry to that of the SDSS data release 12 (DR12). For objects that from our observations are classified as stars, we use SDSS psfMag, while for objects classified as galaxies we use the SDSS magnitude cmodelMag, which is the magnitude from a linear combination of the best-fit exponential and r^1/4 profiles. In all cases, we compare to our standard-calibrated magnitudes mag_auto.

We used two methods in the comparison: (1) a direct comparison of magnitudes of objects in the 10 clusters with available SDSS photometry, and (2) a comparison of star colors to the northern SDSS standard stars (Smith et al. 2002). The second method enables us to evaluate the accuracy of the photometry of all the clusters.

Based on the comparison of SDSS photometry with our photometry calibrated using the initial calibrations (Section 4.1), magnitude zero-point offsets were applied as detailed in Table 10. Figures 10–11 and Table 11 summarize the comparisons after these offsets were applied. Figure 10 shows comparisons for the lowest- and the highest-redshift clusters only, as all other comparisons look similar. The resulting offsets and scatter of the comparisons listed in Table 11 were derived from objects with SDSS magnitude uncertainties less than 0.2 mag, and excluding saturated objects and objects for which our photometry deviates from the SDSS photometry by more than 0.7 mag. The resulting scatter in the comparisons is typically 0.15–0.30 mag, lower for the stars than the galaxies.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Table 10.  Offsets Added to Photometric Zero-points

Cluster	g'	r'	i'	z'	Comments
Abell 1689	0.000	−0.147	−0.117	⋯	SDSS comparison
RX J0056.2+2622 F1	0.000	0.000	0.095	⋯	Internal comparison
RX J1334.3+5030	⋯	0.000	0.000	−0.065	SDSS comparison
RX J0152.7–1357	⋯	0.000	0.000	−0.100	SDSS star colors
RX J1226.9+3332	⋯	−0.139	−0.049	−0.060	SDSS comparison
RXJ 1415.1+3612	⋯	−0.092	−0.159	−0.095	SDSS comparison

Note. Clusters not listed in the table were standard-calibrated with the nominal zero-points.

Download table as:  ASCIITypeset image

Table 11.  Comparison of GMOS Photometry with SDSS Photometry

Field & objects	g' Band			r' Band			i' Band			z' Band
	N	Δ	rms	N	Δ	rms	N	Δ	rms	N	Δ	rms
Abell 1689 galaxiesa	360	−0.013	0.28	405	−0.091	0.26	393	−0.051	0.28	⋯	⋯	⋯
Abell 1689 starsa	62	0.052	0.21	67	0.097	0.19	63	0.051	0.20	⋯	⋯	⋯
RX J0056.2+2622 galaxies	332	−0.049	0.27	400	0.002	0.22	390	0.022	0.22	⋯	⋯	⋯
RX J0056.2+2622 stars	87	−0.021	0.16	98	0.014	0.18	102	0.070	0.14	⋯	⋯	⋯
RX J0142.0+2131 galaxies	99	−0.103	0.24	109	−0.059	0.28	180	0.026	0.21	⋯	⋯	⋯
RX J0142.0+2131 stars	21	0.018	0.12	20	−0.220	0.22	29	0.057	0.13	⋯	⋯	⋯
RX J0027.6+2616 galaxies	91	0.084	0.31	153	0.003	0.24	156	−0.072	0.24	⋯	⋯	⋯
RX J0027.6+2616 stars	34	0.132	0.13	40	0.142	0.15	44	0.020	0.12	⋯	⋯	⋯
Abell 851 galaxies	202	0.042	0.27	302	0.027	0.22	323	−0.003	0.22	⋯	⋯	⋯
Abell 851 stars	27	0.107	0.16	35	0.055	0.17	36	0.043	0.12	⋯	⋯	⋯
RX J2146.0+0423 galaxies	80	0.061	0.28	112	−0.017	0.26	104	0.026	0.23	⋯	⋯	⋯
RX J2146.0+0423 stars	89	0.033	0.17	98	0.015	0.16	98	−0.004	0.14	⋯	⋯	⋯
RX J1334.3+5030 galaxiesa	⋯	⋯	⋯	144	0.047	0.30	178	0.012	0.29	99	0.124	0.29
RX J1334.3+5030 starsa	⋯	⋯	⋯	32	−0.030	0.21	33	−0.006	0.19	25	0.008	0.15
RX J1716.6+6708 galaxies	⋯	⋯	⋯	37	0.087	0.27	48	−0.026	0.27	24	0.178	0.32
RX J1716.6+6708 stars	⋯	⋯	⋯	43	0.059	0.15	45	−0.010	0.10	43	0.053	0.13
RX J1226.9+3332 galaxiesa	⋯	⋯	⋯	100	0.021	0.31	111	0.016	0.28	30	0.308	0.27
RX J1226.9+3332 starsa	⋯	⋯	⋯	39	−0.012	0.15	39	−0.010	0.13	35	0.001	0.12
RX J1415.1+3612 galaxiesa	⋯	⋯	⋯	104	0.042	0.24	101	−0.018	0.26	32	0.128	0.26
RX J1415.1+3612 starsa	⋯	⋯	⋯	39	0.007	0.15	44	0.019	0.19	33	0.065	0.13

Note. Differences are "GMOS"–"SDSS."

^aZero-point offsets were applied to one of more passbands before final comparison; see Table 10.

Download table as:  ASCIITypeset image

The adopted magnitude zero-point offsets (Table 10) represent a compromise between offsets derived from the direct comparisons and achieving colors of the stars in the fields consistent with the locus of the SDSS standard stars in the color–color diagrams shown in Figure 11. In addition, the direct comparisons show systematic differences between the photometry for stars and that for galaxies in the fields. These differences likely originate from differences between methods used in the SDSS and those used in this paper to determine total magnitudes. Based on our simulations presented in Section 3.4, we expect the differences to depend on the distribution of n_ser for the galaxies included in the comparisons. If the comparisons are dominated by disk galaxies, then the differences should be close to zero, while for a mix of disk and bulge-dominated galaxies, the differences are expected to be ≈0.07 mag, increasing to 0.15 mag for a sample of only bulge-dominated galaxies. The median value of the differences between g', r', and i' is −0.03 mag, with 73% of the fields and filters within ±0.07 mag. The median of the differences for four z'-band comparisons is 0.12 mag. For three of the four fields with available SDSS z'-band photometry, we adjusted the magnitude zero-points based primarily on the photometry of the stars in the fields. Thus, it is unlikely that this magnitude offset for the galaxies is related to problem with our zero-points. Finally, the z'-band comparisons for the galaxies show no dependences on the magnitudes, colors, or sizes of the galaxies. For our purpose, we conclude that absolute calibrations of g', r', and i' are consistent with the expected calibration consistency of ≈0.05 mag obtainable with GMOS when using standard methods for calibration; cf. Jørgensen (2009). The z'-band magnitudes may only be consistent to ≈0.12 mag. We discuss this further when establishing the color–magnitude relations for the clusters; see Section 7.

4.3. Calibration Notes for Individual Clusters

Table 12 shows the content of the available machine-readable table of the final calibrated photometric parameters. For each cluster, objects classified as galaxies are listed first, followed by those classified as stars. The columns are as follows:

• 1.  
  Cluster—cluster name.
• 2.  
  ID—GCP ID number for the galaxy.
• 3.  
  and 4. R.A. (J2000), decl. (J2000)—right ascension and declination calibrated to be consistent with USNO (Monet et al. 1998) with an rms scatter of ≈0.7 arcsec.
• 5.  
  to 12. ${g}_{\mathrm{total}}^{{\prime} }$, ${r}_{\mathrm{total}}^{{\prime} }$, ${i}_{\mathrm{total}}^{{\prime} }$, ${z}_{\mathrm{total}}^{{\prime} }$—total magnitude in g', r', i', and z' determined from SExtractor mag_auto and associated uncertainties using the corrections from Table 6.
• 13.  
  to 20. ${g}_{\mathrm{aper}}^{{\prime} }$, ${r}_{\mathrm{aper}}^{{\prime} }$, ${i}_{\mathrm{aper}}^{{\prime} }$, ${z}_{\mathrm{aper}}^{{\prime} }$—aperture magnitudes derived using the aperture sizes defined in Equation (3), and associated uncertainties.
• 21.  
  to 28. ${g}_{2.5}^{{\prime} }$, ${r}_{2.5}^{{\prime} }$, ${i}_{2.5}^{{\prime} }$, ${z}_{2.5}^{{\prime} }$—aperture magnitudes derived using an aperture with diameter 2.5 arcsec, and associated uncertainties.
• 29.  
  P—P(class_star) product of class_star for the available passbands; cf. Equation (4).
• 30.  
  M—M(class_star) median of class_star for the available passbands.

Table 12.  GCP Photometric Catalog

Cluster	ID	R.A. (J2000)	Decl. (J2000)	${g}_{\mathrm{total}}^{{\prime} }$	σg	${r}_{\mathrm{total}}^{{\prime} }$	σr	${i}_{\mathrm{total}}^{{\prime} }$	σi	${z}_{\mathrm{total}}^{{\prime} }$	σz	${g}_{\mathrm{aper}}^{{\prime} }$	σg	${r}_{\mathrm{aper}}^{{\prime} }$	σr	${i}_{\mathrm{aper}}^{{\prime} }$	σi
(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)	(14)	(15)	(16)	(17)	(18)
A1689	1	13:11:12.96	−1:20:31.8	25.234	0.251	24.274	0.191	23.725	0.191	⋯	⋯	⋯	⋯	24.342	0.738	23.666	0.576
A1689	2	13:11:12.99	−1:20:38.7	20.614	0.027	20.106	0.035	19.828	0.038	⋯	⋯	20.770	0.014	20.277	0.018	19.994	0.020
A1689	3	13:11:13.03	−1:19:27.3	23.832	0.162	24.255	0.419	23.021	0.201	⋯	⋯	23.387	0.181	⋯	⋯	22.822	0.240
A1689	4	13:11:13.10	−1:21:45.2	23.356	0.101	22.362	0.079	21.741	0.067	⋯	⋯	23.189	0.128	22.177	0.100	21.510	0.079
A1689	6	13:11:13.16	−1:19:44.1	23.814	0.182	23.800	0.329	22.840	0.202	⋯	⋯	23.624	0.206	23.905	0.484	22.701	0.230
A1689	9	13:11:13.36	−1:21:02.3	23.358	0.053	23.103	0.075	22.494	0.069	⋯	⋯	23.318	0.151	23.116	0.236	22.177	0.144
A1689	13	13:11:13.47	−1:22:05.6	23.260	0.085	22.049	0.057	22.159	0.096	⋯	⋯	23.118	0.117	21.744	0.070	21.964	0.127
A1689	14	13:11:13.54	−1:22:23.9	20.523	0.016	19.151	0.010	18.764	0.010	⋯	⋯	20.584	0.011	19.270	0.007	18.860	0.007
A1689	15	13:11:13.54	−1:19:35.1	18.772	0.028	17.715	0.023	17.174	0.018	⋯	⋯	19.531	0.004	18.353	0.003	17.860	0.003
A1689	16	13:11:13.64	−1:19:41.9	21.918	0.081	21.395	0.095	20.776	0.073	⋯	⋯	22.226	0.054	21.598	0.059	21.101	0.055
${z}_{\mathrm{aper}}^{{\prime} }$	σz	${g}_{2.5}^{{\prime} }$	σg	${r}_{2.5}^{{\prime} }$	σr	${i}_{2.5}^{{\prime} }$	σi	${z}_{2.5}^{{\prime} }$	σz	P	M	log riso	${g}_{\mathrm{iso}}^{{\prime} }$	${r}_{\mathrm{iso}}^{{\prime} }$	${i}_{\mathrm{iso}}^{{\prime} }$	${z}_{\mathrm{iso}}^{{\prime} }$		PA
(19)	(20)	(21)	(22)	(23)	(24)	(25)	(26)	(27)	(28)	(29)	(30)	(31)	(32)	(33)	(34)	(35)	(36)	(37)
⋯	⋯	25.315	0.335	24.170	0.222	23.651	0.221	⋯	⋯	0.00	0.35	−0.498	26.462	25.689	25.366	⋯	0.594	128
⋯	⋯	21.325	0.009	20.840	0.010	20.531	0.013	⋯	⋯	0.00	0.02	0.369	20.694	20.205	19.933	⋯	0.349	171
⋯	⋯	24.202	0.127	24.007	0.188	23.151	0.137	⋯	⋯	0.00	0.36	−0.314	25.139	24.691	24.115	⋯	0.120	164
⋯	⋯	23.632	0.072	22.600	0.052	22.015	0.049	⋯	⋯	0.08	0.55	−0.193	24.028	23.218	22.603	⋯	0.050	24
⋯	⋯	24.164	0.122	23.857	0.165	23.130	0.135	⋯	⋯	0.00	0.24	−0.250	24.886	24.542	23.980	⋯	0.096	163
⋯	⋯	23.349	0.058	23.091	0.082	22.468	0.074	⋯	⋯	0.00	0.12	−0.121	23.503	23.420	22.812	⋯	0.074	19
⋯	⋯	23.382	0.057	22.356	0.043	22.330	0.067	⋯	⋯	0.00	0.08	−0.124	23.593	22.871	22.752	⋯	0.127	30
⋯	⋯	20.900	0.006	19.718	0.004	19.263	0.004	⋯	⋯	0.00	0.03	0.347	20.571	19.250	18.842	⋯	0.106	24
⋯	⋯	19.949	0.003	18.834	0.002	18.311	0.002	⋯	⋯	0.00	0.03	0.735	18.950	17.830	17.329	⋯	0.068	108
⋯	⋯	22.605	0.029	22.057	0.032	21.640	0.035	⋯	⋯	0.00	0.02	0.091	22.562	22.045	21.636	⋯	0.204	134

Note. Columns are explained in Section 5.

Only a portion of this table is shown here to demonstrate its form and content. A machine-readable version of the full table is available.

Download table as:  DataTypeset image

For those objects classified as galaxies, the table also contains the following columns:

• 31.  
  log r_iso—the logarithm (base 10) of the isophotal circularized radius in arcseconds in the detection band; cf. Equation (7). See Table 5 for the surface brightness limit at the isophote.
• 32.  
  to 35. ${g}_{\mathrm{iso}}^{{\prime} }$, ${r}_{\mathrm{iso}}^{{\prime} }$, ${i}_{\mathrm{iso}}^{{\prime} }$, ${z}_{\mathrm{iso}}^{{\prime} }$—isophotal magnitudes derived using the surface brightness limits listed in Table 5.
• 36.  
  —ellipticity in the detection band, as derived from the semimajor and semiminor axes using SExtractor,  = 1 − a/b.
• 37.  
  PA—position angle in the detection band in degrees measured north through east.

All magnitudes and colors in the table are AB magnitudes. Magnitude measurements with uncertainties larger than 1 mag are omitted from the table. Galactic extinction for each field and filter are listed in Table 16 in Appendix A. The data in Table 12 have not been corrected for Galactic extinction.

Uncertainties are included for total magnitudes and colors. Uncertainties on the isophotal magnitudes are similar to those on the total magnitudes. The typical uncertainties on the logarithm of the isophotal radii are 0.003 with the largest uncertainties 0.01–0.015. For completeness, we list the isophotal radii even when they are smaller than the seeing of the image in the detection filter. The uncertainties on the ellipticities are typically of similar size to the magnitude uncertainties in the detection band. The ellipticities have not been corrected for the effect of the image quality, thus they are expected to be affected by systematic errors, especially for galaxies smaller than about twice the seeing of the images.

Based on internal comparisons, we evaluate that the uncertainties on the position angles are <3° for galaxies with  ≥ 0.3 and total magnitude in the detection band of 23 mag or brighter. Uncertainties are <5° for galaxies with  = 0.1–0.3 and total magnitude in the detection band of 22 mag or brighter. Position angles of fainter or less elliptical galaxies are subject to higher uncertainties.

We calibrate the photometry, total magnitudes, and colors to the rest frames of the galaxies using calibrations based on stellar population models from Bruzual & Charlot (2003). We first described our method in Jørgensen et al. (2005). Here we generalize the method to calibrate the total magnitudes to the rest-frame B band for all clusters and also establish the calibration of the colors to rest-frame (U – B) and (B – V). The rest-frame B magnitudes, (U – B) and (B – V) used here are Vega magnitudes.

We use single stellar population (SSP) models from Bruzual & Charlot for a Chabrier (2003) initial mass function, ages of 2–13 Gyr, metallicities of Z = 0.004, 0.008, 0.02 (solar), and 0.04, and Padova 1994 evolutionary tracks; see Bruzual & Charlot for original references to the tracks. In our previous calibrations (Jørgensen et al. 2005; Jørgensen & Chiboucas 2013), we used filter functions included in the software distributed by Bruzual & Charlot. The filter functions for the SDSS filters g', r', i', and z' are identical to those supplied by the SDSS. Thus, we maintained use of these filter functions. However, the filter functions for U, B, and V are from Buser & Kurucz (1978). Filter functions for these filters were shown by Maíz Apellániz (2006) to give inaccurate descriptions of data. Maíz Apellániz derived better filter functions and also eliminated the internally inconsistent use of two filter functions for the B filter. We have here adopted these newer filter functions for U, B, and V. Below we comment on the effect of this compared to our previously used calibrations.

The Bruzual & Charlot SSP models were used to derive rest-frame B, (U – B), and (B – V) (Vega magnitudes), as well as the observed AB magnitudes g', r', i', and z', and colors. This was done in steps of 0.025 in redshift and for the redshift range spanning our observations. For each of these redshifts, we established the calibration to rest-frame B as

$$\begin{eqnarray}&&{B}_{\mathrm{rest}}={m}_{\mathrm{obs}}+{\alpha }_{1}\cdot {\mathrm{color}}_{\mathrm{obs},1}+{\beta }_{1},\end{eqnarray} \tag{ 11 }$$

where m_obs is the magnitude in the observed band most closely matching the rest-frame B at the redshift and color_obs is the observed color best complementing m_obs to achieve coverage of the full rest-frame B band. Inclusion of a second-order color term, as we did in Jørgensen et al. (2005) and Jørgensen & Chiboucas (2013), does not significantly improve the calibrations, when using the Maíz Apellániz (2006) U, B, and V filter functions. From B_rest, the absolute B-band magnitude is derived as

$$\begin{eqnarray}&&{M}_{B}={B}_{\mathrm{rest}}-\mathrm{DM}(z),\end{eqnarray} \tag{ 12 }$$

where DM(z) is the distance modulus for a given redshift. Similarly, we establish calibrations to rest-frame (U – B) and (B – V) at each of the redshifts as

$$\begin{eqnarray}&&(U-B)={\alpha }_{2}\cdot {\mathrm{color}}_{\mathrm{obs},2}+{\beta }_{2}\end{eqnarray} \tag{ 13 }$$

and

$$\begin{eqnarray}&&(B-V)={\alpha }_{3}\cdot {\mathrm{color}}_{\mathrm{obs},3}+{\beta }_{3},\end{eqnarray} \tag{ 14 }$$

where color_obs,2 and color_obs,3 are the observed colors from the passbands most closely matching the passbands for rest-frame (U – B) and (B – V), respectively.

The calibrations to rest-frame B, (U – B), and (B – V) are applied to the data by interpolating the calibration coefficients to the exact redshift of each of the galaxies. It is important to note that the validity of the calibrations do not rely on the models being successful at modeling the ages and metallicities of the stellar populations in the observed galaxies. Rather, the models only have to provide correct relative color information over the wavelength range spanned by the desired rest-frame passbands and the observed passbands used in the calibration. As long as extrapolations from the observed passbands to the desired rest-frame passbands are kept to a minimum and the available models do span the observed colors, any shortcomings of the models to reproduce the exact colors of galaxies for physically believable ages and metallicities are less important. Additional information on how to calibrate the photometry to a "fixed-frame" system, i.e., rest-frame B, can be found in Blanton et al. (2003).

Figure 12 shows our previous B-band calibration compared with the one established here using the filter functions from Maíz Apellániz (2006). The calibrations are shown at the model redshifts closest to the redshifts of the three clusters analyzed in Jørgensen & Chiboucas (2013). The difference between two calibrations is typically 0.05 mag in rest-frame B magnitudes, with the new calibration leading to fainter magnitudes. This change has no significant effect on our previously published results. Future analysis of the GCP data will use the calibrations established in the present paper. Figure 13 shows the color calibrations for three typical redshifts spanning the GCP cluster sample, demonstrating that linear calibrations are sufficient to fit the model data and provide reliable calibrations.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

In Table 13 we provide the calibrations matching the cluster redshifts, as well as the distance moduli for the clusters. For guidance on how the optimal calibrations were chosen, we also list the effective wavelength of each of the observed bands in the cluster rest frames. In most cases, the observed colors used in the calibrations match the optimal redshift intervals, except for RX J2146.0+0423 for which no z' imaging was obtained. For the highest-redshift clusters, the calibrations to (B – V) in all cases rely on the same photometry as the calibrations to (U – B) and B. Thus, they rely on extrapolation of the data using the SSP models.

Table 13.  Rest-frame Calibrations at the Cluster Redshifts

Cluster	Redshift	λeff in Cluster Rest Framea				DM(z)	Brest	(U – B)	(B – V)
		g'	r'	i'	z'
Reference	0.0000	475	630	780	925
Abell 1689	0.1865	400	531	657	⋯	39.78	$g^{\prime} +0.336-0.440(g^{\prime} -r^{\prime} )$	$1.005(g^{\prime} -r^{\prime} )-0.801$	$0.673(g^{\prime} -r^{\prime} )+0.079$
							± 0.007 ± 0.006	± 0.028 ± 0.034	± 0.007 ± 0.008
RX J0056.2+2622	0.1922	398	528	654	⋯	39.86	g' + 0.343 − 0.461(g' − r')	0.989 (g' − r') − 0.800	0.662 (g' − r') + 0.079
							± 0.008 ± 0.006	± 0.027 ± 0.033	± 0.007 ± 0.009
RX J0142.0+2131	0.2794	371	492	610	⋯	40.78	r' + 0.434 + 0.297(g' − r')	0.852 (g' − r') − 0.845	1.460 (r'−i') + 0.127
							± 0.011 ± 0.008	± 0.015 ± 0.021	± 0.020 ± 0.011
RX J0027.6+2616	0.3650	348	462	571	⋯	41.45	$r^{\prime} +0.508+0.123(g^{\prime} -r^{\prime} )$	0.798 (g' − r') − 0.919	1.300 (r'−i') + 0.127
							± 0.007 ± 0.005	± 0.004 ± 0.006	± 0.005 ± 0.003
Abell 851	0.4050	338	448	555	⋯	41.72	$r^{\prime} +0.574+0.054(r^{\prime} -i^{\prime} )$	$0.867(g^{\prime} -r^{\prime} )-1.035$	$1.163(r^{\prime} -i^{\prime} )+0.139$
							± 0.001 ± 0.005	± 0.005 ± 0.008	± 0.004 ± 0.002
RX J1347.5–1145	0.4506	327	434	538	⋯	41.99	$r^{\prime} +0.609-0.211(r^{\prime} -i^{\prime} )$	$0.576(g^{\prime} -i^{\prime} )-0.975$	$0.992(r^{\prime} -i^{\prime} )+0.141$
							± 0.001 ± 0.001	± 0.007 ± 0.017	± 0.006 ± 0.004
RX J2146.0+0423	0.532	310	411	509	⋯	42.42	$r^{\prime} +0.677-0.553(r^{\prime} -i^{\prime} )$	$0.541(g^{\prime} -i^{\prime} )-1.000$	$0.751(r^{\prime} -i^{\prime} )+0.148$
							± 0.005 ± 0.005	± 0.010 ± 0.026	± 0.007 ± 0.007
MS 0451.6–0305	0.5398	308	409	507	601	42.46	$i^{\prime} +0.683+0.422(r^{\prime} -i^{\prime} )$	$1.103(r^{\prime} -i^{\prime} )-0.700$	$0.739(r^{\prime} -i^{\prime} )+0.146$
							± 0.005 ± 0.005	± 0.033 ± 0.033	± 0.008 ± 0.008
RX J0216.5–1747	0.578	301	399	494	586	42.64	$i^{\prime} +0.706+0.322(r^{\prime} -i^{\prime} )$	$1.029(r^{\prime} -i^{\prime} )-0.709$	$1.513(i^{\prime} -z^{\prime} )+0.151$
							± 0.007 ± 0.006	± 0.024 ± 0.026	± 0.019 ± 0.009
RX J1334.3+5030	0.620	⋯	389	481	571	42.83	$i^{\prime} +0.710+0.237(r^{\prime} -i^{\prime} )$	$0.985(r^{\prime} -i^{\prime} )-0.741$	$1.500(i^{\prime} -z^{\prime} )+0.122$
							± 0.007 ± 0.006	± 0.016 ± 0.018	± 0.011 ± 0.006
RX J1716.6+6708	0.809	⋯	348	431	511	43.53	$i^{\prime} +0.865-0.442(i^{\prime} -z^{\prime} )$	$0.690(r^{\prime} -z^{\prime} )-0.973$	$0.854(i^{\prime} -z^{\prime} )+0.169$
							± 0.002 ± 0.003	± 0.004 ± 0.009	± 0.005 ± 0.004
MS 1610.4+6616	0.8300	⋯	344	426	505	43.60	i'+0.879 − 0.513(i' − z')	$0.681(r^{\prime} -z^{\prime} )-0.976$	$0.790(i^{\prime} -z^{\prime} )+0.174$
							± 0.003 ± 0.003	± 0.004 ± 0.009	± 0.005 ± 0.005
RX J0152.7–1357	0.8350	⋯	343	425	504	43.62	$i^{\prime} +0.881-0.528(i^{\prime} -z^{\prime} )$	$0.678(r^{\prime} -z^{\prime} )-0.975$	$0.780(i^{\prime} -z^{\prime} )+0.175$
							± 0.003 ± 0.003	± 0.005 ± 0.010	± 0.006 ± 0.005
RX J1226.9+3332	0.8908	⋯	333	413	489	43.79	$i^{\prime} +0.901-0.665(i^{\prime} -z^{\prime} )$	$0.652(r^{\prime} -z^{\prime} )-0.965$	$0.706(i^{\prime} -z^{\prime} )+0.165$
							± 0.005 ± 0.005	± 0.006 ± 0.013	± 0.010 ± 0.010
RX J1415.1+3612	1.0269	⋯	311	385	456	44.17	$z^{\prime} +0.981-0.024(i^{\prime} -z^{\prime} )$	$0.610(r^{\prime} -z^{\prime} )-0.980$	$0.688(i^{\prime} -z^{\prime} )+0.076$
							± 0.002 ± 0.002	± 0.012 ± 0.026	± 0.025 ± 0.030

Note. The second line for each cluster lists the uncertainties on the calibration coefficients.

^aWavelengths noted only for the passbands that were obtained for each of the clusters.

Download table as:  ASCIITypeset image

The spectroscopic samples for the GCP were selected based on magnitudes and colors, and when available at the time of sample selection, redshift information from the literature. The aim was to include the maximum number of galaxies on the red sequence. Our previous papers describe the sample selection for the clusters for which we have published the spectroscopic data; see Barr et al. (2005) for RX J0142.0+2131, Jørgensen et al. (2005) for RX J0152.7–1357, Jørgensen & Chiboucas (2013) for MS 0451.6–0305 and RX J1226.9+3332, and Jørgensen et al. (2017) for Abell 1689, RX J0056.2+2622, RX J0027.6+2616, and RX J1347.5–1145. These papers also include grayscale images of the clusters with the spectroscopic samples labeled. For the remainder of the clusters, we summarize the sample selection in Table 17 in Appendix B, and in Appendix C provide grayscale images with the spectroscopic samples marked, Figures 20–27. The grayscale image of RX J0142.0+2131 is included in the present paper, as the X-ray data were not available at the time of publication of our previous paper on the cluster (Barr et al. 2005).

Figure 14 shows the color–magnitude relations for the clusters, using the colors for the B-band rest-frame calibration. For galaxies on the red sequence, the figure shows both aperture colors from aperture sizes defined in Equation (3) and total colors. We fit the color–magnitude relations, using aperture colors, for the members iteratively (red symbols on Figure 14), rejecting galaxies deviating by more than three times the scatter relative to the relation. The rejection was iterated four times to reach a stable fit of the red sequence. The best fits to the cluster members are shown as red solid lines and summarized in Table 14. For clusters for which our spectroscopic data have not yet been processed, we instead fit the relations to the spectroscopic sample, excluding those galaxies with blue colors in at least one of the available colors. The fits for these clusters were also determined iteratively with rejection. The difference between using aperture colors and colors based on mag_auto has a median of 0.03 on the zero-points (total colors being bluer), with an rms scatter of 0.05. The slopes and scatter of the relations are not significantly different for the two sets of colors. Thus, we proceed using only the aperture colors.

Zoom In Zoom Out Reset image size

Table 14.  Color–Magnitude Relations in the Observed Frame

Cluster	Relations	rms	N
Abell 1689	$(g^{\prime} -r^{\prime} )=(-0.027\pm 0.009)(g^{\prime} -19)+(1.233\pm 0.010)$	0.059	61
	$(r^{\prime} -i^{\prime} )=(-0.015\pm 0.005)(g^{\prime} -19)+(0.476\pm 0.006)$	0.032	64
RX J0056.2+2622	$(g^{\prime} -r^{\prime} )=(-0.034\pm 0.008)(g^{\prime} -19)+(1.161\pm 0.013)$	0.063	52
	$(r^{\prime} -i^{\prime} )=(-0.010\pm 0.004)(g^{\prime} -19)+(0.412\pm 0.005)$	0.027	51
RX J0142.0+2131	$(g^{\prime} -r^{\prime} )=(-0.046\pm 0.015)(r^{\prime} -20)+(1.374\pm 0.017)$	0.074	24
	$(r^{\prime} -i^{\prime} )=(-0.016\pm 0.005)(r^{\prime} -20)+(0.549\pm 0.006)$	0.027	25
RX J0027.6+2616	$(g^{\prime} -r^{\prime} )=(-0.062\pm 0.011)(r^{\prime} -20)+(1.750\pm 0.009)$	0.046	28
	$(r^{\prime} -i^{\prime} )=(-0.008\pm 0.005)(r^{\prime} -20)+(0.650\pm 0.004)$	0.022	30
Abell 851	$(g^{\prime} -r^{\prime} )=(0.001\pm 0.027)(r^{\prime} -20)+(1.718\pm 0.027)$	0.137	41
	$(r^{\prime} -i^{\prime} )=(0.007\pm 0.013)(r^{\prime} -20)+(0.575\pm 0.013)$	0.070	44
RX J1347.5–1145	$(g^{\prime} -r^{\prime} )=(-0.028\pm 0.025)(r^{\prime} -20)+(1.540\pm 0.041)$	0.136	41
	$(r^{\prime} -i^{\prime} )=(-0.001\pm 0.007)(r^{\prime} -20)+(0.734\pm 0.011)$	0.035	40
RX J2146.0+0423	$(g^{\prime} -r^{\prime} )=(-0.063\pm 0.015)(r^{\prime} -21)+(1.725\pm 0.020)$	0.076	20
	$(r^{\prime} -i^{\prime} )=(-0.014\pm 0.014)(r^{\prime} -21)+(0.894\pm 0.019)$	0.073	20
MS 0451.6–0305	$(g^{\prime} -r^{\prime} )=(-0.059\pm 0.031)(r^{\prime} -21)+(1.686\pm 0.024)$	0.121	39
	$(r^{\prime} -i^{\prime} )=(-0.027\pm 0.012)(r^{\prime} -21)+(0.892\pm 0.009)$	0.047	36
	$(i^{\prime} -z^{\prime} )=(-0.001\pm 0.009)(r^{\prime} -21)+(0.381\pm 0.007)$	0.036	42
RX J0216.5–1747	$(g^{\prime} -r^{\prime} )=(-0.115\pm 0.032)(i^{\prime} -21)+(1.813\pm 0.023)$	0.118	30
	$(r^{\prime} -i^{\prime} )=(-0.057\pm 0.013)(i^{\prime} -21)+(1.021\pm 0.010)$	0.048	28
	$(i^{\prime} -z^{\prime} )=(-0.071\pm 0.011)(i^{\prime} -21)+(0.386\pm 0.008)$	0.043	33
RX J1334.3+5030	$(r^{\prime} -i^{\prime} )=(-0.005\pm 0.015)(i^{\prime} -22)+(1.115\pm 0.027)$	0.050	18
	$(i^{\prime} -z^{\prime} )=(0.020\pm 0.012)(i^{\prime} -22)+(0.441\pm 0.022)$	0.033	18
RX J1716.6+6708	$(r^{\prime} -i^{\prime} )=(-0.002\pm 0.021)(i^{\prime} -22)+(1.271\pm 0.020)$	0.060	27
	$(i^{\prime} -z^{\prime} )=(-0.002\pm 0.010)(i^{\prime} -22)+(0.645\pm 0.009)$	0.027	24
MS 1610.4+6616a	$(r^{\prime} -i^{\prime} )=1.329\pm 0.019$	0.038	4
	$(i^{\prime} -z^{\prime} )=0.588\pm 0.022$	0.043	4
RX J0152.7–1357	$(r^{\prime} -i^{\prime} )=(-0.027\pm 0.026)(i^{\prime} -22)+(1.358\pm 0.022)$	0.075	27
	$(i^{\prime} -z^{\prime} )=(0.009\pm 0.005)(i^{\prime} -22)+(0.698\pm 0.004)$	0.013	25
RX J1226.9+3332	$(r^{\prime} -i^{\prime} )=(-0.012\pm 0.016)(i^{\prime} -22)+(1.141\pm 0.013)$	0.085	49
	$(i^{\prime} -z^{\prime} )=(0.000\pm 0.007)(i^{\prime} -22)+(0.702\pm 0.006)$	0.039	50
RX J1415.1+3612	$(r^{\prime} -i^{\prime} )=(0.000\pm 0.042)(i^{\prime} -22)+(1.097\pm 0.028)$	0.086	14
	$(i^{\prime} -z^{\prime} )=(-0.040\pm 0.011)(i^{\prime} -22)+(0.773\pm 0.009)$	0.030	16

Note.

^aMedian colors of the four passive galaxies in the z = 0.83 group.

Download table as:  ASCIITypeset image

We then evaluate the completeness of the spectroscopic samples along the red sequence, including galaxies within ±3 times the scatter for the color–magnitude relations as marked on Figure 14. Figure 15 shows the distribution of the absolute B-band magnitudes, M_B,abs, of the spectroscopic samples, together with the completeness. In general, the samples are at least 80% complete for galaxies brighter than M_B,abs ≤ −22 mag, except when the spatial distribution of these galaxies made it impossible to include all of them in the mask designs for the spectroscopic observations. This was the case for RX J0027.6+2616 and RX J1226.9+3332. For −22 < M_B,abs ≤ −19.5 mag, the samples for clusters at z < 0.5 typically include 20%–50% of the red sequence galaxies. For higher-redshift clusters, the samples are limited at M_B,abs ≈ −20 mag, but reach the same completeness.

Zoom In Zoom Out Reset image size

While the main purpose of this paper is to present the consistently calibrated X-ray measurements for the GCP clusters and the full photometric catalog from the optical imaging, we take this opportunity to briefly discuss the changes in the color–magnitude relations as a function of redshift.

Figure 16 shows the observed mean colors of the red sequence as a function of cluster redshift. The colors are the zero-points listed for the color–magnitude relations in Table 14 and correspond to the colors at M_B,abs ≈ −21 mag. Models from Bruzual & Charlot (2003) are overlaid for ages of 2.5, 5, and 10 Gyr, and solar metallicity [M/H] = 0. The (g' − r') and (r' − i') colors follow the expected variation with redshift, consistent with mean ages of 2.5–10 Gyr and solar metallicity. The (i' − z') colors are systematically bluer than predicted by the models. Figure 16(c) includes a low-metallicity model with [M/H] = −0.7 and age = 2.5 Gyr. Comparing with the models from Vazdekis et al. (2012) and available from the MILES Web site, we find that the MILES models for a Chabrier IMF and the BaSTI isochrones are 0.05–0.10 bluer in (i' − z') than in the Bruzual & Charlot models. Further, Mei et al. (2009) find the red sequence of RX J0152.7–1357 to have (r₆₂₅–z₈₅₀) = 1.93 at i₇₇₅ = 22.5 in AB magnitudes. We find (r' − z') = 2.04 at i' = 22.5 for this cluster. While the two photometric systems are not completely identical, we take the comparison as an indication that it is unlikely that our colors are significantly too blue. Our z'-band mag_auto for the galaxies may be too faint at the 0.12 mag level; cf., Section 4.2. However, the (i' − z') aperture colors for the stars are consistent with SDSS; see Figure 11. In conclusion, it is possible that the Bruzual & Charlot (2003) models do not correctly model the z'-band magnitudes.

Zoom In Zoom Out Reset image size

To further investigate the changes in the color–magnitude relations with redshift, we establish the relations based on the absolute B-band magnitudes, and (U – B) and (B – V) colors in the rest frames of the clusters. Figures 17–18 and Table 15 summarize the slopes, zero points, and scatter of the color–magnitude relations. The fits are based on confirmed members for those clusters with fully processed spectroscopy. For clusters without processed spectroscopy, the fits are based on the same selection of spectroscopic samples as used for establishing the color–magnitude relations in the observed frames; cf., Section 7. The unusually positive slope for the RX J1334.3+5030 (B – V)–magnitude relation may be a result of the inclusion of faint non-members. However, the zero point at M_B,abs = −21 appears to be affected by less than 0.05 mag, so we make no attempt here to exclude additional galaxies from the fit. In Figure 19 we show the colors at M_B,abs = −21, the slopes, and internal scatter of the relations as a function of cluster redshift. The internal scatter was derived by subtracting off in quadrature the median uncertainty at the observed magnitude corresponding to M_B,abs = −21. We do not include any contribution from the calibration to the rest frame, as random errors from the observations dominate over random errors from the rest-frame calibration. The figure also includes data from Cerulo et al. (2016), Foltz et al. (2015), and Mei et al. (2009) covering redshifts from 0.8 to 1.5. The literature data have been calibrated to also show colors at M_B,abs = −21 (Vega magnitudes) and slopes of the relations relative to the absolute B-band magnitude.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

In Figures 19(a) and (b), passive evolution models based on Bruzual & Charlot (2003) are shown for formation redshifts of z_form = 1.4–4.0 and solar metallicity. For z_form = 1.8, we also show low- and high-metallicity models to illustrate how the assumed metallicity affects the predicted colors. Our results are generally in agreement with passive evolution and a formation redshift of 1.5–2.0, consistent with our results based on the absorption line indices (Jørgensen et al. 2017). At redshifts z = 0.8–1.0, where our coverage overlaps with that in Cerulo et al. (2016), our (U – B) results are in agreement, while our (B – V) colors are ≈0.15 bluer than that found by Cerulo et al. We caution that our wavelength coverage for the highest-redshift clusters does not overlap with the V band and therefore the (B – V) colors rely on extrapolation based on the Bruzual & Charlot models. The (U – B) colors from Mei et al. and Foltz et al. are ≈0.1–0.15 redder than our results and those from Cerulo et al. In general, the zero-points of the color–magnitude relations provide a much less stringent constraint on the ages of intermediate-redshift cluster galaxies than the absorption line indices (e.g., Jørgensen & Chiboucas 2013; Jørgensen et al. 2017). However, the results still serve as a consistency check of the results.

We find no change in the slope of the color–magnitude relations as a function of redshift; see Figures 19(c) and (d). This is in agreement with results from Cerulo et al. (2016), Foltz et al. (2015), and Mei et al. (2009). Thomas et al. (2005) established age–velocity dispersion and metallicity–velocity dispersion relations at z ≈ 0. We use those relations and the Faber-Jackson relation (Faber & Jackson 1976; luminosity–velocity dispersion relation) established for the joint sample of Abell 1689 and RX J0056.2+2622 members to derive the expected slopes of the color–magnitude relations, under the assumption of passive evolution. The predictions are shown as the dashed lines in Figures 19(c) and (d). Assuming no age variation with velocity dispersion and adopting the metallicity–velocity dispersion relation from Thomas et al. give predicted slopes of the color–magnitude relations indicated by the dotted–dashed lines in Figures 19(c) and (d). The joint data from the GCP (this paper), Cerulo et al. (2016), Foltz et al. (2015), and Mei et al. (2009) are inconsistent with the low-redshift age–velocity dispersion relation seen simply as a consequence of passive evolution. This limits the allowable age change along the color–magnitude relation from the brightest cluster galaxies to galaxies four magnitudes fainter to <0.05 dex. Alternatively, a steep slope of the low-redshift age–velocity dispersion relation must be maintained by adding younger galaxies to the red sequence, possibly primarily at low masses; cf., McDermid et al. (2015).

We also find no change in the internal scatter of the color–magnitude relations as a function of redshift; see Figures 19(e) and (f). One might expect that the addition of younger galaxies to the red sequence at later epochs would lead to a higher scatter at lower redshifts. However, the samples used for establishing the color–magnitude relations are incomplete at low luminosities and also biased against galaxies far from the red sequence, as they are simply our spectroscopic samples aimed at galaxies on the red sequence. In addition, the transition from blue star-forming galaxy to passive red galaxy may be too fast to result in significantly higher scatter. Only Abell 851 in the GCP sample contains a significant number of post-starburst bulge-dominated galaxies (P. Hibon et al. 2018, in preparation). The color–magnitude relations for this cluster does have a significantly higher scatter than found for the other GCP clusters. The results from Cerulo et al. (2016), Foltz et al. (2015), and Mei et al. (2009) are based on larger photometric samples. However, these authors also use sigma clipping when fitting the color–magnitude relations, most likely affecting the estimates of the scatter.

In this paper, we have given an overview of the science goals for the Gemini/HST Galaxy Cluster Project (GCP), summarized the cluster selection, and assembled consistently calibrated X-ray measurements for the clusters. We present the photometric catalog based on the ground-based imaging of the GCP clusters in the g', r', i', and z' bands. The photometry has been calibrated to be consistent with the SDSS photometric system. The sample selection for the spectroscopic observations are summarized and provided for those clusters not included in prior publications.

We established the calibration of the photometry to rest-frame magnitudes and colors, and provide calibration coefficients at the relevant cluster redshifts.

Finally, we have derived the color–magnitude relations for all clusters and briefly discussed the redshift dependence on the red sequence mean color and scatter, and compare our results to results from the literature for higher-redshift clusters and to stellar population models. The absence of change in the slopes of the rest-frame color–magnitude relations with redshifts limits the allowable age differences along the color–magnitude relation to <0.05 dex from the brightest cluster galaxies to those four magnitudes fainter. The data add evidence to the need for younger, low-mass galaxies to be added to the red sequence between z ≈ 1 and the present in order to obtain a relatively steep age–velocity dispersion relation at low redshift.

The Gemini TACs and the former Gemini Director Matt Mountain are thanked for generous time allocations to carry out the Gemini/HST Galaxy Cluster Project. The anonymous referee is thanked for comments and suggestions that helped improve this paper. The following collaborators are thanked for their contributions to the proposals, obtaining the observations, and/or data processing during the early phases of the project: Jordi Barr, Marcel Bergmann, Maela Collobert, David Crampton, Roger Davies, and Bryan Miller. We thank Harald Ebeling for confirming ahead of publication that RX J1415.1+3612 was a z ≈ 1 cluster suitable for inclusion in the GCP. Pierluigi Cerulo is thanked for comments on the rest-frame calibration and the redshift evolution of the color–magnitude relations.

Based on observations obtained at the Gemini Observatory, which is operated by the Association of Universities for Research in Astronomy, Inc., under a cooperative agreement with the NSF on behalf of the Gemini partnership: the National Science Foundation (United States), the National Research Council (Canada), CONICYT (Chile), Ministério da Ciência, Tecnologia e Inovação (Brazil) and Ministerio de Ciencia, Tecnología e Innovación Productiva (Argentina). The data were obtained while also the Science and Technology Facilities Council (United Kingdom), and the Australian Research Council (Australia) contributed to the Gemini Observatory.

The data presented in this paper originate from the following Gemini programs: GN-2001B-Q-10, GN-2001B-DD-3, GN-2002A-Q-34, GN-2002B-Q-29, GN-2002B-DD-4, GN-2003A-DD-4, GN-2003B-Q-21, GN-2003B-DD-3, GS-2003B-Q-26, GN-2004A-Q-45, and GS-2005A-Q-27, System Verification program GN-2001B-SV-51, and engineering programs GN-2002B-SV-90 and GN-2003A-SV-80. The data were processed using the Gemini IRAF package.

Photometry from SDSS is used for comparison purposes. Funding for the SDSS and SDSS-II has been provided by the Alfred P. Sloan Foundation, the Participating Institutions, the National Science Foundation, the U.S. Department of Energy, the National Aeronautics and Space Administration, the Japanese Monbukagakusho, the Max Planck Society, and the Higher Education Funding Council for England. The SDSS web site is http://www.sdss.org/.

Observations that were obtained with XMM-Newton, an ESA science mission funded by ESA member states and NASA; the Chandra X-ray Observatory; and the ROSAT Mission have been used. The observations were obtained from these missions' data archives.

Table 16 gives detailed information on the available observations, exposure times, image quality, and sky brightness. The table also lists the adopted Galactic extinction for each field and filter.

The sample selection for the GCP spectroscopic observations is based on the color–magnitude diagrams and any redshift information available in the literature at the time of sample selection. Table 17 summarizes the sample selection for those clusters not included in our previous papers. Objects in classes 1 and 2 have highest priority, and span the brighter 2–2.5 magnitude of the red sequence. Available redshifts were used to give higher priority to members and if possible excluded non-members from this selection. We used redshifts from Gioia et al. (1999) for RX J1716.6+6708 and from Dressler et al. (1999) for Abell 851. Objects in class 3 are typically fainter galaxies on the red sequence. However, for clusters with some prior redshift information, brighter red sequence galaxies without prior redshift determinations are included in class 3. Objects in class 4 are only added to fill available space in the mask design. They are typically fainter and/or bluer than the objects in classes 1–3. For the field of MS 1610.4+6616, the sample selection was aimed at the published cluster redshift of 0.65 (Luppino & Gioia 1995). Objects in class 1 were assigned to targets with colors that could match this redshift and with a limit of i' = 21.8 mag. However, because the field does not contain a rich cluster, there is no well-defined red sequence at the expected colors. Redder and bluer, or fainter, objects were assigned object class 2, and given lower priority in the mask design. The redder objects turned out to be part of the poor group that we identified at z = 0.83. The sample in RX J1334.3+5030 was selected without a blue limit on the colors and using only two intervals of the total magnitude to ensure coverage in luminosities. Thus, the sample contains a much larger fraction of blue galaxies than the samples in the other fields.

Table 17.  Selection Criteria for Spectroscopic Samples

Cluster	Obj. Class	Selection criteria
Abell 851a	1	Confirmed member based on redshift ∧ 18.4 ≤ r' ≤ 20.9 ∧ 0.3 ≤ (r' − i') ≤ 0.8
	2	Confirmed member based on redshift ∧ 20.9 < r' ≤ 21.7 ∧ 0.3 ≤ (r' − i') ≤ 0.8
	3	No redshift ∧ 18.4 ≤ r' ≤ 21.7 ∧ 0.3 ≤ (r' − i') ≤ 0.8
	4	[18.4 ≤ r' ≤ 21.7 ∧ ((r' − i') < 0.3 (r' − i') > 0.8)]   21.7 < r' ≤ 22.7
RX J2146.0+0423	1	i' ≤ 20.55 ∧ (g' − r') ≥ 1.0  ∧ (r' − i') ≥ 0.55  ∧ (g' − r') ≥ 0.4 + 1.3(r' − i')
	2	20.55 < i' ≤ 21.65 ∧ (g' − r') ≥ 1.0  ∧ (r' − i') ≥ 0.55  ∧ (g' − r') ≥ 0.4 + 1.3(r' − i')
	3	21.65 < i' ≤ 22.55 ∧ (g' − r') ≥ 1.0  ∧ (r' − i') ≥ 0.55  ∧ (g' − r') ≥ 0.4 + 1.3(r' − i')
	4	[23. < i' ≤ 22.55 ∧ (g' − r') ≥ 1.0 ∧ (r' − i') ≥ 0.55 ∧ (g' − r') ≥ 0.4 + 1.3(r' − i')]
		[i' ≤ 23 ∧ ((g' − r') < 1.0 (r' − i') < 0.55  (g' − r') < 0.4 + 1.3(r' − i'))]
RX J0216.5–1747b	1	i' ≤ 20.5 ∧ (g' − r') ≥ 1.2  ∧ (r' − i') ≥ 0.7
	2	20.5 < i' ≤ 21.6 ∧ (g' − r') ≥ 1.2  ∧ (r' − i') ≥ 0.7
	3	21.6 < i' ≤ 22.6 ∧ (g' − r') ≥ 1.2  ∧ (r' − i') ≥ 0.7
	4	i' ≤ 22.6 ∧ [(g' − r') < 1.2 (r' − i') < 0.7]
RX J1334.3+5030a	1	i' ≤ 20.7 ∧ (r' − i') ≤ 1.35
	2	20.7 < i' ≤ 22.4 ∧ (r' − i') ≤ 1.35
RX J1716.6+6708	1	Confirmed member based on redshift ∧ i' ≤ 21.6 ∧ 0.9 ≤ (r' − i') ≤ 1.5  ∧ 0.45 ≤ (i' − z') ≤ 0.75
	2	[Confirmed member based on redshift ∧ 21.6 < i' ≤ 22.5 ∧ 0.9 ≤ (r' − i') ≤ 1.5 ∧ 0.45 ≤ (i' − z') ≤ 0.75]
		[No redshift ∧ i' ≤ 21.6 ∧ 0.9 ≤ (r' − i') ≤ 1.5 ∧ 0.45 ≤ (i' − z') ≤ 0.75]
	3	No redshift ∧ 21.6 < i' ≤ 22.5 ∧ 0.9 ≤ (r' − i') ≤ 1.5  ∧ 0.45 ≤ (i' − z') ≤ 0.75
	4	i' ≤ 22.5 ∧ [(r' − i') < 0.9 (i' − z') < 0.45]
MS 1610.4+6616	1	i' ≤ 21.8 ∧ 0.8 ≤ (r' − i') ≤ 1.2  ∧ 0.3 ≤ (i' − z') ≤ 0.45
	2	i' < 23 ∧ ![ 0.8 ≤ (r' − i') ≤ 1.2 ∧ 0.3 ≤ (i' − z') ≤ 0.45]
RX J1415.1+3612	1	i' ≤ 22 ∧ 0.6 ≤ (i' − z') ≤ 0.8  ∧ (r' − i') ≤ 1.3
	2	22 < i' ≤ 22.5 ∧ 0.6 ≤ (i' − z') ≤ 0.8  ∧ (r' − i') ≤ 1.3
	3	22.5 < i' ≤ 23.5 ∧ 0.6 ≤ (i' − z') ≤ 0.8  ∧ (r' − i') ≤ 1.3
	4	i' ≤ 23.5 ∧ (i' − z') < 0.6

Notes.

^aAt the time of sample selection, only imaging in r' and i' was available. ^bAt the time of sample selection, only imaging in g', r', and i' was available.

Download table as:  ASCIITypeset image

Here we summarize the properties of the clusters, with reference to the original discovery papers, results regarding cluster structure (evidence for subclusters or merging), and the star formation history.

In the descriptions, we make use of Figures 20–27, showing the grayscale images of the clusters for which such information was not included in our previous papers. The images include overlaid contours of X-ray data from either XMM-Newton or Chandra, or in the case of the field MS 1610.4+6616, X-ray data from ROSAT. We also show the grayscale image for RX J0142.0+2131 since at the time of the original publication (Barr et al. 2005), the Chandra X-ray data were not available. Graycale images of the remaining GCP clusters are available in Jørgensen et al. (2005, RX J0152.7–1357), Jørgensen & Chiboucas (2013, MS 0451.6–0305, RX J1226.9+3332), and Jørgensen et al. (2017, Abell 1689, RX J0056.2+2622, RX J0027.6+2626, RX J1347.5–1145).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

For the clusters RX J0142.0+2131, Abell 851, RX J0216.5–1747, and RX J1415.1+3612, our spectroscopic samples are marked with information about cluster membership. For MS 1610.4+6616, we show the spectroscopic sample with the 12 members of the poor group indicated. The processing of the spectroscopic data for RX J2146.0+0423, RX J1334.3+5060, and RX J1716.6+6708 is pending. Thus, for these clusters, we show the spectroscopic sample divided into galaxies on the red sequence and those bluer than the red sequence. The labeling matches our selection for the fits to the red sequence; see Section 7.

Abell 1689/RX J1311.4–0120. This cluster is included in the Abell catalog of northern clusters (Abell et al. 1989), and has been observed with XMM-Newton and Chandra; see Jørgensen et al. (2017) for the X-ray data overlaid on our imaging data. The cluster velocity dispersion is very high, ≈2100 km s⁻¹ (Jørgensen et al. 2017; Czoske 2004). Analysis of the cluster kinematic data and lensing data (Lemze et al. 2009; Umetsu et al. 2015) shows that the central structure is complex and that the X-ray mass estimate is likely too low. The cluster may be a merger; see discussion in Andersson & Madejski (2004). The cluster is included in our spectroscopic analysis of GCP data (Jørgensen et al. 2017) where we find that the stellar populations of the bulge-dominated galaxies are consistent with the median metallicities and abundance ratios for the GCP clusters.

RX J0056.2+2622/Abell 115. This cluster is also included in the Abell catalog of northern clusters (Abell et al. 1989). This is a binary cluster; see the grayscale image in our analysis in the paper Jørgensen et al. (2017). Barrena et al. (2007) found that the two subclusters are in the plane of the sky as based on the kinematic structure of the cluster. The northern subcluster brightest galaxy is the powerful radio galaxy 3C 28 and hosts an active galactic nucleus (AGN); see, e.g., Hardcastle et al. (2009). The galaxy is ID 1054 in our spectroscopic sample. The cluster has been observed with XMM-Newton and Chandra. The cluster X-ray emission is quite diffuse, showing the presence of the two subclusters; see the overlay of XMM-Newton data on our optical imaging in Jørgensen et al. (2017). We find that the stellar populations of the bulge-dominated galaxies in the cluster are consistent with the median metallicities and abundance ratios for the GCP clusters (Jørgensen et al. 2017).

RX J0142.0+2131. This cluster is included in the northern ROSAT All-Sky Galaxy Cluster Survey (NORAS; Böhringer et al. 2000) and the extended ROSAT Brightest Cluster Sample (eBCS; Ebeling et al. 2000a). A bright foreground galaxy is superimposed on the cluster near the center. This led Böhringer et al. to mistakenly list the cluster redshift as 0.0696, the redshift of the foreground galaxy, although the correct cluster redshift is z = 0.28. At the time of the publication of our spectroscopic study of this cluster, Barr et al. (2005), XMM-Newton and Chandra data were not available. Chandra data have since been obtained. Figure 20 shows the X-ray data overlaid on our optical imaging. The morphological appearance is that of a relaxed cluster, with X-ray point sources associated with optical counterparts. No detailed analysis of the X-ray data seems to be available in the literature.

RX J0027.6+2616. This cluster was discovered in ROSAT observations and first listed in NORAS by Böhringer et al. (2000) and also included in the eBCS (Ebeling et al. 2000a). Later observations with Chandra show little substructure, except for possibly some X-ray emission associated with members of the foreground group at z = 0.34, which we identified from our optical spectroscopy Jørgensen et al. (2017). The full spectroscopic analysis is included in Jørgensen et al.

Abell 851. This cluster is included in the Abell catalog of northern clusters (Abell et al. 1989). The cluster is very massive and contains substantial substructure. Based on XMM-Newton observations, De Filippis et al. (2003) identified two main subclusters with internal structure; see also Figure 21 for the X-ray data overlaid on our optical imaging. This cluster contains a large fraction of disk galaxies as well as post-starburst galaxies (Andreon et al. 1997; Oemler et al. 2009), and as such is quite atypical for a massive cluster at this redshift. Andreon et al. find the spiral fraction to be close to 50% and hypothesize that the reason may be that the relatively low density intercluster gas failed to stop the star formation in the cluster members. Oemler et al. focus on starburst and post-starburst galaxies in the cluster, and in particular find that the youngest starburst galaxies reside in the center of the cluster. Our spectral analysis of the GCP data will be presented in P. Hibon et al. (2018, in preparation).

RX J1347.5–1147. This cluster was discovered as the most X-ray luminous of the ROSAT clusters (Schindler et al. 1997). It has been studied extensively in both the optical and X-ray. Weak- and strong-lensing studies have been conducted in an attempt to better estimate the cluster mass and understand the dynamical structure of the cluster; see, e.g., Bradač et al. (2008). As discussed in Jørgensen et al. (2017), there is evidence that the cluster contains an infalling substructure to the southeast of the cluster center (Ettori et al. 2004; Kreisch et al. 2016). We also found that the velocity dispersion of the cluster is in agreement with expectations from the X-ray luminosity, once corrected for the diffuse emission from the substructure. The cluster is included in the analysis in Jørgensen et al.

RX J2146.0+0423. The cluster was first mentioned by Gunn et al. (1986) in their photographic survey for intermediate-redshift clusters. The cluster appeared in the 160 square degree ROSAT survey (Vikhlinin et al. 1998), with the cluster redshift listed by Mullis et al. (2003). The cluster was also included in the Wide Angle ROSAT Pointed Survey (WARPS; Perlman et al. 2002). Figure 22 shows our optical image of the cluster overlaid with the X-ray data from XMM-Newton. The cluster is one of the lowest-mass clusters included in the GCP and appears relatively compact, with the majority of the red galaxies in our spectroscopic sample within one arcminute of the cluster center. Our spectroscopic analysis of the cluster will be presented in a future paper.

MS 0451.6–0305. This cluster was the most X-ray-luminous cluster included in the Einstein Extended Medium Sensitive Survey (EMSS; Gioia & Luppino 1994). Based on ROSAT data, it was estimated to be among the most X-ray-luminous clusters above redshift 0.5 (Ebeling et al. 2007). The CNOC survey found a very large cluster velocity dispersion, 1330 ± 100 km s⁻¹, confirming the large mass of the cluster (Ellingson et al. 1998; Borgani et al. 1999). This is in agreement with our result for the velocity dispersion, ${1450}_{-159}^{+100}\,\mathrm{km}\,{{\rm{s}}}^{-1}$ (Jørgensen & Chiboucas 2013). Strong-lensing modeling by Zitrin et al. (2011) shows that the central mass distribution may be elliptical, while weak-lensing studies show that the brightest cluster galaxy is slightly offset from the peak of the X-ray emission (Comerford et al. 2010; Hoekstra et al. 2012; Soucail et al. 2015). Thus, the cluster is most likely not relaxed. Moran et al. (2007a, 2007b) used wide-field HST/ACS data and optical spectroscopy to study the morphological evolution and star formation history. These authors concluded that the star formation history is truncated, and that star formation stopped at an epoch corresponding to a formation redshift of ≈2. The cluster is included in our analysis in Jørgensen & Chiboucas (2013) and Jørgensen et al. (2017). Based on the absorption line strengths, we find that the bulge-dominated galaxies in this cluster on average have ≈0.1 dex lower metallicity than found for other GCP clusters.

RX J0216.5–1747. The cluster was included in WARPS (Perlman et al. 2002). Together with RX J2146.0+0423, RX J1334.3+5030, and RX J1716.6+6708, this cluster is among the four lowest-mass clusters in the GCP. Figure 23 shows our optical image of the cluster overlaid with the X-ray data from Chandra. The cluster appears less compact than RX J2146.0+0423 and RX J1716.6+6708. Our spectroscopic analysis of the cluster will be presented in a future paper.

RX J1334.3+5030. This cluster was included in the Bright Serendipitous High-Redshift Archival Cluster (SHARC) survey (Romer et al. 2000). The cluster is characterized as non-relaxed by Parekh et al. (2015) based on the morphology of the X-ray emission. Figure 24 shows our optical image of the cluster overlaid with the X-ray data from XMM-Newton. Our spectroscopic analysis of the cluster will be presented in a future paper.

RX J1716.6+6708. This cluster was discovered as part of the ROSAT North Ecliptic Pole Survey (Henry et al. 1997). At the time of discovery, this cluster was among only four known z > 0.75 X-ray clusters. Gioia et al. (1999) found from 37 member galaxies a cluster velocity dispersion of ≈1500 km s⁻¹, which is significantly higher than expected given the X-ray luminosity of the cluster. Figure 25 shows our optical image of the cluster overlaid with the X-ray data from Chandra. The X-ray morphology shows some substructure possibly associated with two concentrations of red galaxies about 1.5 arcminute from the otherwise compact core of the cluster. The cluster was included, together with the two higher-redshift GCP clusters RX J0152.7–1357 and RX J1226.9+3332, in the investigation of the K-band luminosity function for z = 0.6–1.3 clusters by De Propris et al. (2007), who found the luminosity functions to be consistent with massive galaxies being fully in place by z ≈ 1.3, but also that the epoch of major star formation was as recent as z = 1.5–2. Our spectroscopic analysis of the cluster will be presented in a future paper.

MS 1610.4+6616. The EMSS (Gioia & Luppino 1994) included MS 1610.4+6616 as a cluster at redshift larger than 0.5, while Luppino & Gioia (1995) gives a redshift of 0.65. However, subsequent observations have shown that this is not a rich cluster. Donahue et al. (1999) state that the X-ray source is a point source and possibly originates from an AGN. Our spectroscopic observations confirm that the field does not contain a rich cluster. Our sample contains 27 galaxies with redshifts of z = 0.60–0.86, 12 of which are clustered at z = 0.83. However, there is no clear clustering of these galaxies and no obvious extended X-ray emission associated with them; see Figure 26. For completeness, we include the photometry obtained of this field in the present paper.

RX J0152.7–1357. This cluster was originally discovered from ROSAT data and is included in three different ROSAT surveys: the ROSAT Deep Cluster Survey, the Bright SHARC survey (Nichol et al. 1999), and WARPS (Ebeling et al. 2000b). The XMM-Newton and Chandra X-ray observations show that the cluster consists of two subclusters, probably in the process of merging (Maughan et al. 2003). See also Jørgensen et al. (2005) for the XMM-Newton X-ray data overlaid on our optical imaging. Nantais et al. (2013) studied the morphologies of the member galaxies and put forward the hypothesis that infalling galaxies are transformed directly from peculiar systems into bulge-dominated galaxies. The cluster is included in our analysis in Jørgensen et al. (2005, 2006, 2007, 2017) and Jørgensen & Chiboucas (2013). The steep slope of the FP for the cluster relative to that of our low-redshift reference sample supports that low-mass bulge-dominated galaxies contain younger stellar populations than the higher-mass galaxies. This is also supported by the analysis of the spectroscopic and photometric data presented by Demarco et al. (2010). In addition, we find that the average of the abundance ratios [α/Fe] derived from our spectra are ≈0.2 dex higher than that of the other GCP clusters.

RX J1226.9+3332. This cluster was discovered in WARPS (Ebeling et al. 2001). The X-ray structure is due to AGNs. However, Maughan et al. (2007) analyzed the temperature map of the X-ray gas and concluded that it showed evidence of a recent merger event, which is associated with the overdensity of the galaxies southwest of the cluster center; see the grayscale image of the cluster in our analysis paper, Jørgensen & Chiboucas (2013). Recent analysis based on the Sunyaev-Zel'dovich effect of the cluster supports this view (Adam et al. 2015). The cluster is included in our analysis in Jørgensen et al. (2006, 2007, 2017) and Jørgensen & Chiboucas (2013). The FP for this cluster is consistent with our results for RX J0152.7–1357 and indicates the presence of stellar populations in the lower-mass galaxies that are younger than those in higher-mass galaxies.

RX J1415.1+3612. The highest-redshift cluster in our z = 0.2–1.0 GCP sample was first listed in WARPS (Perlman et al. 2002), with the note added in proof giving the spectroscopic confirmation of z = 1.013 for the brightest cluster galaxy. Huang et al. (2009) studied strong lensing created by the cluster. These authors find the cluster redshift to be z = 1.026 and cluster velocity dispersion of σ_cl = 807 ± 185 km s⁻¹, in agreement with our measurements; cf., Table 1. Figure 27 shows our optical image of the cluster overlaid with the X-ray data from Chandra. Our spectroscopic analysis of the cluster will be presented in a future paper.