Spectroscopy of the Neighboring Massive Clusters Abell 222 and Abell 223 Night for the Masks Used During That Night, Before the Science Exposures Were Made

Astronomy & Astrophysics manuscript no. h3727 October 25, 2018 (DOI: will be inserted by hand later) Spectroscopy of the neighboring massive clusters Abell 222 and Abell 223 ⋆ J. P. Dietrich1, D. I. Clowe1, and G. Soucail2 1 Institut für Astrophysik und Extraterrestrische Forschung, Universität Bonn, Auf dem Hügel 71, 53121, Bonn, Germany 2 Observatoire Midi–Pyrénées, UMR5572, 14 Av. Edouard´ Belin, 31400 Toulouse, France Received 29.05.2002 / Accepted 09.08.2002 Abstract. We present a spectroscopic catalog of the neighboring massive clusters Abell 222 and Abell 223. The catalog contains the positions, redshifts, R magnitudes, V −R color, as well as the equivalent widths for a number of lines for 183 galaxies, 153 of them belonging to the A 222 and A 223 system. We determine the heliocentric redshifts to be z = 0.2126 ± 0.0008 for A 222 and z = 0.2079 ± 0.0008 for A 223. The velocity dispersions of +90 −1 +99 −1 both clusters in the cluster restframe are about the same: σ = 1014 −71 km s and σ = 1032 −76 km s for A 222 and A 223, respectively. While we find evidence for substructure in the spatial distribution of A 223, no kinematic substructure can be detected. From the red cluster sequence identified in a color–magnitude–diagram we determine the luminosity of both clusters and derive mass–to–light ratios in the R–band of (M/L)A222 = (202±43) h70 M⊙/L⊙ and (M/L)A223 = (149±33) h70 M⊙/L⊙. Additionally we identify a group of background galaxies at z ∼ 0.242. Key words. Galaxies: clusters: general – Galaxies: clusters: individual: A 222 – Galaxies: clusters: individual: A 223 – Galaxies: distances and redshifts – Galaxies: luminosity function, mass function 1. Introduction data and discuss deviations from previous values in the literature. The spatial distribution and the kinematics of A 222/223 are two Abell clusters at z ≈ 0.21 separated ′ −1 the double cluster system are examined in Sect. 3 with an by ∼ 14 on the sky, or ∼ 2600h70 kpc, belonging to emphasis on finding possible substructure. We determine the Butcher et al. (1983) photometric sample. Both clus- the luminosity and mass–to-light ratio of the clusters by ters are rich having Abell richness class 3 (Abell 1958). selecting the red cluster sequence in Sect. 4. Our results While these are optically selected clusters, they have been are summarized in Sect. 5. Throughout this paper we as- observed by ROSAT (Wang & Ulmer 1997; David et al. −1 −1 sume an ΩΛ =0.7, Ωm =0.3, H0 = 70 h70 km s Mpc arXiv:astro-ph/0208266v1 14 Aug 2002 1999) and are confirmed to be massive clusters. 9 spectra cosmology. of galaxies in the cluster region, most of them being cluster members, were known (Sandage et al. 1976; Newberry et al. 1988) before Proust et al. (2000, hereafter PEL) 2. Data and Data Reduction published a list of 53 spectra and did a first kinematical study of this system. PEL also found 4 galaxies at the clus- Multi-object spectroscopy of the two clusters Abell 222 ter redshift in the region between the clusters (hereafter and Abell 223 was performed at the NTT on three con- “intercluster region”), indicating a possible connection be- secutive nights in December 1999. These nights were tween the clusters. clear with occasional high cirrus. The instrument used was EMMI with grism 2, which has a resolution of 580 We report 184 independent redshifts for 183 galaxies at 600 nm and a dispersion of 11.6 nm/mm. With the in the field of Abell 222 and Abell 223, more than three 2048x2048 CCD pixels of 24 µm this leads to a dispersion times the number of redshifts previously known, as well of 0.28 nm/pixel. With one exception two exposures of as equivalent widths for a number of lines. 2700 seconds each were taken for 6 fields, 3 on each cluster. The paper is organized as follows. In Sect. 2 we de- For the field centered on A 222 in the second night only scribe the reduction of the spectroscopic and photometric one exposure of 2700 seconds was available. The wave- Send offprint requests to: J. P. Dietrich, e-mail: length calibration was done using Helium–Argon lamps, [email protected] which provided typically 20 lines used in the calibration. ⋆ Based on observations made at ESO – La Silla (Chile) The calibration frames were taken at the beginning of the 2 J. P. Dietrich et al.: Spectroscopy of the neighboring massive clusters Abell 222 and Abell 223 night for the masks used during that night, before the science exposures were made. 0.25 2.1. Reduction of spectroscopic data For the data reduction a semi-automated IRAF1 package was written by the authors that cuts out the single spectra of the CCD frames and then processes these spectra using standard IRAF routines for single slit spectroscopy. The 0.2 sky spectrum was removed from all spectra using a linear 0.22 fit with a 2σ rejection on measurements on each side of the galaxy spectrum where the position of the galaxy on 0.215 the slit permitted it. Measurements from only one side of the spectrum were used otherwise. A 2σ rejection was 0.21 used to remove cosmic rays and hot pixels. Remaining hot 0.15 pixels or cosmic rays in the sky spectrum introduced fake 0.205 absorption features, while hot pixels or cosmic rays in the 0.2 spectrum itself lead to fake emission features. These were 0.2 0.205 0.21 0.215 0.22 removed by hand. 0.15 0.2 0.25 Because the sky spectrum removal was done column z (this work) by column and no distortion correction was applied, some Fig. 1. Comparison of redshift measurements for objects residual sky lines remained in the final spectra, most no- observed by us and with redshifts listed in PEL. The large tably of the strong [O i] emission at 5577 A.˚ At the typical panel shows the full sample, the inset is a blow up of the redshift of the cluster members of z ≈ 0.21 this line does cluster region. The error bars are the internal errors re- not coincide with any important feature and thus does not ported by RVSAO. cause any problems in the subsequent analysis. 2.2. Redshift determination their and our cluster members, indicating that PEL ob- The radial velocity determination was carried out using served a sub–sample with a significantly different mean the cross-correlation method (Tonry & Davis 1979) im- value from the larger sample we describe here. plemented in the RVSAO package (Kurtz & Mink 1998). We ruled out the possibility that this discrepancy could Spectra of late type stars and elliptical galaxies with have been caused by taking all calibration frames before known radial velocities were used as templates. The red- the science exposures were taken. This could have intro- shift determination was verified by visual inspection of duced a shift of the zero point if the masks were not moved identified absorption and emission features. The average back to their original position for the science exposures. internal error reported by RVSAO is cz = 68 km s−1. We confirmed that this is not the case by determining Adding this in quadrature to the estimated error of the the radial velocity of the subtracted sky spectrum in the wavelength calibration of cz = 110 km s−1, we estimate wavelength calibrated frames. We found that, if a zero −1 the error in the redshift determination of single galaxies point shift occurred, it must be smaller than 30 km s , to be δz =0.0004. confirming the accuracy of our data. Fig. 1 shows a comparison of our redshift measurements and the redshifts listed by PEL. The obvious out- 2.3. Equivalent widths lier in the large panel is from the sample of Sandage et al. (1976). The inset shows a broad agreement between our We measured equivalent widths for the [O ii]λ3727, results and the values of PEL. Ignoring the obvious out- [O iii]λ5007 emission lines, and Hβ and Hα emission and lier, the average difference between the measurements is absorption lines. The integration ranges for the features −3 hz − zPELi = (0.2±1.1)×10 . Student’s t-test rejects the and the continuum were fixed by the values given in Table null hypothesis of different sample means with same vari- 1. ance at higher than the 99% level for the 23 cluster galax- [O ii] and Hα are important indicators of star forma- ies we have in common with PEL. However, Student’s tion rates (Kennicutt 1998). To accurately determine the t-test confirms the hypothesis of different sample means equivalent widths and in particular estimate their signif- with same variance at higher than the 99% level for all icance level we follow the definition of equivalent widths given by Czoske et al. (2001): 1 IRAF is distributed by the National Optical Astronomy Observatories, which are operated by the Association of Nint Universities for Research in Astronomy, Inc., under cooper- fi Wλ = ∆λ − Nint∆λ, (1) ative agreement with the National Science Foundation. i=1 fc X J. P. Dietrich et al.: Spectroscopy of the neighboring massive clusters Abell 222 and Abell 223 3 Table 1. Restframe wavelength ranges for equivalent widths measurement. All wavelengths are given in A.˚ The last column gives the continuum range that was used for estimating the signal–to–noise ratio. Feature λcent line blue cont.

Load more