An extension of the Planck galaxy cluster catalogue
R. A. Burenin (IKI, Moscow)

TL;DR
This paper extends the Planck galaxy cluster catalogue by identifying about 3000 clusters using combined Planck, WISE, and SDSS data, significantly increasing known clusters at higher redshifts for future surveys.
Contribution
It provides an expanded galaxy cluster catalogue with improved completeness and higher redshift coverage, aiding future large-scale structure studies.
Findings
Approximately 3000 clusters identified in SDSS fields.
Catalogue is highly complete for clusters with M_500 > 3x10^14 Msun.
Contains about ten times more clusters than previous Planck SZ catalogues at z>0.4.
Abstract
We present a catalogue of galaxy clusters detected in the Planck all-sky Compton parameter maps and identified using data from the WISE and SDSS surveys. The catalogue comprises about 3000 clusters in the SDSS fields. We expect the completeness of this catalogue to be high for clusters with masses larger than M_500 =~ 3x10^14 Msun, located at redshifts z<0.7. At redshifts above z=~0.4, the catalogue contains approximately an order of magnitude more clusters than the 2nd Planck Catalogue of Sunyaev-Zeldovich sources in the same fields of the sky. This catalogue can be used for identification of massive galaxy clusters in future large cluster surveys, such as the SRG/eROSITA all-sky X-ray survey.
| Number | S/N | Note | ||||||
|---|---|---|---|---|---|---|---|---|
| (J2000) | arc min | |||||||
| 1 | 4.58 | |||||||
| 2 | 7.77 | PSZ2 G107.67-39.78 | ||||||
| 3 | 4.55 | |||||||
| 4 | 9.52 | PSZ2 G104.30-48.99 | ||||||
| 5 | 4.99 | ∗∗ | ||||||
| 6 | 5.86 | |||||||
| 7 | 14.20 | ∗∗ PSZ2 G092.16-66.01 | ||||||
| 8 | 6.10 | |||||||
| 9 | 7.21 | ∗∗ PSZ2 G099.57-58.64 | ||||||
| 10 | 5.52 | |||||||
| 11 | 4.87 | |||||||
| 12 | 5.36 | |||||||
| 13 | 7.22 | |||||||
| 14 | 5.19 | ∗∗ | ||||||
| 15 | 11.20 | PSZ2 G105.40-50.43 | ||||||
| 16 | 5.64 | |||||||
| 17 | 4.83 | |||||||
| 18 | 6.80 | |||||||
| 19 | 8.91 | PSZ2 G101.55-59.03 | ||||||
| 20 | 4.68 | ∗∗ | ||||||
| 21 | 5.42 | |||||||
| 22 | 4.83 | ∗∗ | ||||||
| 23 | 10.04 | PSZ2 G104.71-54.54 | ||||||
| 24 | 4.72 | |||||||
| 25 | 8.18 | |||||||
| 26 | 7.46 | |||||||
| 27 | 7.46 | ∗∗ | ||||||
| 28 | 5.65 | ∗∗ | ||||||
| 29 | 8.31 | PSZ2 G109.22-44.01 | ||||||
| 30 | 8.59 | PSZ2 G104.98-54.79 | ||||||
| 31 | 6.37 | |||||||
| 32 | 9.68 | PSZ2 G112.35-32.86 | ||||||
| 33 | 5.83 | ∗∗ | ||||||
| 34 | 4.65 | |||||||
| 35 | 16.12 | PSZ2 G113.29-29.69 | ||||||
| 36 | 9.62 | PSZ2 G108.71-47.75 | ||||||
| 37 | 5.07 | |||||||
| 38 | 5.26 | |||||||
| 39 | 8.86 | ∗∗ PSZ2 G094.46-69.65 | ||||||
| 40 | 6.03 | |||||||
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
An extension of the Planck galaxy cluster catalogue
R. A. Burenin
Space Research Institute RAS, Moscow
Abstract
We present a catalogue of galaxy clusters detected in the Planck all-sky Compton parameter maps and identified using data from the WISE and SDSS surveys. The catalogue comprises about 3000 clusters in the SDSS fields. We expect the completeness of this catalogue to be high for clusters with masses larger than , located at redshifts . At redshifts above , the catalogue contains approximately an order of magnitude more clusters than the 2nd Planck Catalogue of Sunyaev-Zeldovich sources in the same fields of the sky. This catalogue can be used for identification of massive galaxy clusters in future large cluster surveys, such as the SRG/eROSITA all-sky X-ray survey.
keywords:
galaxy clusters, sky surveys
††journal: ApJS††journal: \astl††journal: \astl††journal: MNRAS††journal: MNRAS††journal: MNRAS††journal: ApJ††journal: AJ††journal: ApJS††journal: AJ††journal: ApJS††journal: ApJS††journal: ApJ††journal: A&A††journal: \astl††journal: AJ††journal: AJ††journal: ApJ††journal: ApJ††journal: ApJS††journal: A&A††journal: A&A††journal: A&A††journal: A&A††journal: A&A††journal: A&A††journal: A&A††journal: A&A††journal: A&A††journal: A&A††journal: MNRAS††journal: ApJ††journal: ApJ††journal: ApJ††journal: ApJ††journal: \astl††journal: AJ
\journalinfo
2017001[0]
\submitted
20.11.2016
1 Introduction
Measurement of the galaxy cluster mass function is one of the most sensitive methods used to constrain the parameters of the cosmological model (e.g., Vikhlinin et al., 2009a, b; Planck Collaboration, 2014a, 2016b). For such studies, large samples of massive galaxy clusters are needed.
One of the largest samples of massive galaxy clusters is the catalogue of clusters detected via the Sunyaev-Zeldovich (SZ) effect (Sunyaev, Zeldovich, 1972) in the Planck all-sky survey (Planck Collaboration, 2014b; Planck Collaboration, 2016c). In this survey, the most massive clusters in the observable Universe are detected nearly uniformly over the entire extragalactic sky. The 2nd Planck Catalogue of SZ sources (PSZ2, Planck Collaboration, 2016c) contains 1653 objects, of which 1203 are confirmed massive galaxy clusters. Most of these clusters have masses larger than , i.e. they are the most massive clusters in the Universe. The number density of such objects is very small and their mass function is very steep.
Since the amplitude of the SZ effect depends mostly on galaxy cluster mass, lowering the detection limit in the Planck SZ survey should enable finding objects of lower mass, which would lead to a rapid increase in the number of detected clusters. For example, with a two times lower detection limit, , the number of detected clusters is expected to increase by an order of magnitude (e.g., Vikhlinin et al., 2009b). The cluster detection limit could be lowered if it were possible to use additional data for identification of clusters and elimination of false detections.
Below, we demonstrate that a useful sample of galaxy clusters with masses above at redshifts can be obtained using the Planck all-sky Compton parameter maps in combination with data from the Wide-Field Infrared Survey Explorer (WISE) and Sloan Digital Sky Survey (SDSS). We present a catalogue of about 3000 galaxy clusters found using these data in the SDSS fields.
2 Source detection in Planck Compton parameter maps
The detection of SZ sources in the Planck Collaboration catalogues was done using specialized procedures that take the spectral and spatial shape of the source into account (see, e.g., Planck Collaboration, 2016c). In addition, Compton parameter maps (-maps) were constructed (Planck Collaboration, 2016a), which were mainly used for studying the angular power spectrum of the SZ signal. It has been demonstrated that there is a good agreement between the objects from the SZ source catalogue and the sources detected in the -maps. Therefore, for simplicity, we have used the Planck Compton parameter maps to detect galaxy clusters.
We exploited maps of the Compton parameter and its standard deviation from the Planck 2015 data release (Planck Collaboration, 2016a), as provided by the Planck Legacy Archive111http://pla.esac.esa.int/. We performed our source search using the NILC maps, because they have somewhat lower noise at small angular scales (see details in Planck Collaboration, 2016a). We smoothed the standard deviation map with a -radius median filter. We then obtained a signal to noise map from the -map and the smoothed standard deviation map and additionally subtracted large scale anisotropy, which was estimated by smoothing the map with a -radius median filter.
As a Galaxy foreground mask we used a mask produced by Khatri (2016). This mask makes allowance not only for Galaxy dust emission but also for the emission of carbon monoxide (CO), in particular in high-latitudes molecular clouds. The spectrum of the CO signal resembles the spectrum of -distortions, so that CO emission significantly contaminates the Planck -maps. Specifically, we used the 61% CO mask available in the public domain222http://theory.tifr.res.in/khatri/szresults/.
Figure 1 shows the -parameter signal to noise ratio distribution obtained as discussed above inside the Galactic foreground mask. Small deviations of the -parameter from the mean value are well described by a Gaussian distribution, since the observed signal consists of instrumental noise and the averaged signal of a large number of faint SZ sources. At signal to noise ratios approximately , the distribution differs significantly from the Gaussian one. Obviously, individual SZ sources start to appear above the noise level at these deviations from the mean value.
Since we use additional IR and optical data for cluster identification, which allows us to effectively eliminate false detections from the sample, we can use a lower source detection threshold to detect sources in the -parameter map. We decided to use a source detection threshold (shown with the dotted line in Fig. 1). Although a lot of projections of faint SZ sources may appear near this threshold, we expect to find also many real individual massive clusters, which can be identified using IR and optical data.
We considered only SZ sources inside the foreground mask discussed above. Also, only sources at Galactic latitudes were considered, otherwise source identification in optical and IR bands would be difficult due to strong contamination from Galactic stars. We have thus selected 20290 SZ sources, of which 9227 are located in the SDSS fields. This number of detected sources indicates that the cluster mass threshold in our sample has been lowered not more than two times compared to the 2nd Planck Catalogue of SZ sources. Therefore, these sources (apart from the nearest ones) are not expected to be identified with clusters less massive than .
3 Identification of clusters in the optical and infrared
3.1 The redMaPPer catalogue
Galaxy clusters can be efficiently found using optical photometry, since most galaxies in clusters have similar colours and form the so-called red sequence in the colour-magnitude diagram (e.g., Gladders, Yee, 2000). To identify SZ sources detected in the Planck Compton parameter map, we used a galaxy cluster catalogue obtained from SDSS data using the redMaPPer cluster detection algorithm. Specifically, we used the publicly available redMaPPer catalogue, version 6.3333http://risa.stanford.edu/redMaPPer/, which contains about 26000 galaxy clusters, all having relatively good photometric redshift and richness estimates. Cluster richness is known to correlate well with total cluster mass.
Since we are interested in massive galaxy clusters, with masses above , we considered only clusters with , where is the cluster richness provided in the redMaPPer catalogue. This restriction allows us to reject low mass clusters but to keep nearly all clusters with masses included in the reference sample (Rozo et al., 2015).
Figure 2 shows (red dashed line) the fraction of confirmed clusters from the PSZ2 catalogue that are detected in the SDSS fields using the redMaPPer algorithm (according to the version of the catalogue used here). We have applied the richness limit discussed above, but this has almost no effect on the number of identified clusters, as expected. Note that the redMaPPer mask has not been taken into account here, which may be one of the main reasons why the maximum identified fraction is less than 100%. We also see that the fraction of detected clusters drops at redshifts .
3.2 Clusters identified using WISE and SDSS data
In order to more efficiently identify clusters at higher redshifts, we used IR data from the WISE survey in addition to SDSS optical data. The WISE all-sky survey (Wright et al., 2010) started in 2009 and was initially done in four photometric bands: 3.4, 4.6, 12 and 22 m. Since the end of the cryogenic phase in 2010, the survey has been continuing in the 3.4 and 4.6 m bands (Mainzer et al., 2014). For galaxy cluster observations, the 3.4 m photometric band is most useful. In this band, distant galaxy clusters are well detected at redshifts up to – (e.g., Burenin, 2015).
To search for galaxy clusters in the 3.4 m band images of the WISE all-sky survey, we used a completely automated algorithm that builds on the procedure described in our previous paper (Burenin, 2015). To detect clusters, we first subtracted stars from the WISE images, then detected extended IR sources in these images by convolving them with -models of various angular sizes, and finally identified the brightest cluster galaxies and red sequences inside the detected IR sources using SDSS photometric data.
As compared to our earlier work cited above, the following improvements were made. We used more recent coadds of WISE and NEOWISE images, presented in Meisner et al. (2016) and available for public use444http://unwise.me/ (see also, Lang, 2014). Flux measurements for the sources in WISE images were made using a more complete PSF model, taking into account not only its wings at large angular scales but also its angular asymmetry relative to its center. The data on source positions were taken from SDSS, data release 13 (SDSS Collaboration, 2017). Source flux fitting was done with frozen source positions in the sky, i.e. using so-called “forced photometry”. For brighter galaxies, where the form of the galaxy should be taken in account in addition to the PSF model, we used the “forced photometry” from Lang et al. (2016). In addition, we improved the red sequence detection procedure. The current version of this cluster detection algorithm is a preliminary one. We plan to further improve it in the future, but even this preliminary version is suitable for identification of massive galaxy clusters among Planck SZ sources.
Since only massive clusters can appear among the SZ sources from the Planck survey, we should look for clusters with masses above in WISE images. In order to estimate galaxy cluster masses from IR data, cluster IR luminosities can be used (e.g., Lin et al., 2004; Kopylova, Kopylov, 2006). Figure 3 shows the correlation between the IR luminosity555Hereafter, the cosmological model with , and km с*-1* Mpc*-1* is adopted. in the 3.4 m band and the total mass for clusters in the PSZ2 catalogue. The IR luminosity was calculated as the combined luminosity of red sequence galaxies within the 1 Mpc projected radius. The zero point and bandwidth calibrations were adopted from Jarrett et al. (2011). The K-corrections in the 3.4 m band were calculated using the 11 Gyr age synthetic stellar population model template taken from Bruzual, Charlot (2003), which appears to be a suitable model for all red sequence colours for clusters at redshifts .
We see from Fig. 3 that the cluster IR luminosity estimated in this way correlates well with the total cluster mass. The power law slope of this correlation is . The slope is smaller than unity since the luminosity is calculated within a constant physical radius, rather than within a radius of constant density contrast. The scatter of the IR luminosity near this correlation, excluding larger than deviations, is or . Therefore, IR luminosities allow us to obtain total cluster mass estimates that are nearly as accurate as mass estimates obtained from cluster X-ray luminosities (Vikhlinin et al., 2009b) or optical richness (Rozo et al., 2015).
To identify clusters among Planck SZ sources, we used only clusters with IR luminosities L_{3.4\mbox{\scriptsize\mum}}>2\times 10^{44} erg s*-1*. We see from Fig. 3 that most of the clusters with masses have IR luminosities above this limit. The fraction of confirmed clusters from the PSZ2 catalogue in the SDSS fields that are detected with our automated procedure, with the above IR luminosity constraint applied, is shown in Fig. 2 by the solid blue line. We see that nearly 90% of PSZ2 clusters at are detected using WISE data, while this fraction drops rapidly at . Therefore, we have managed to significantly increase the completeness of cluster identification at redshifts , as compared to the redMaPPer catalogue.
Figure 4 compares photometric redshift estimates obtained with our automated cluster detection procedure with spectroscopic redshifts. These estimates were obtained using the cluster red sequence colours and the magnitudes of cluster brightest galaxies from SDSS and WISE photometric data. To calibrate our photometric estimates, we used clusters from the PSZ2 catalogue (Planck Collaboration, 2016c) as well as from the 400 square degree X-ray cluster survey (400d, Burenin et al., 2007) and the earlier 160 square degree cluster survey (160d, Vikhlinin et al., 1998) based on ROSAT pointings data. To increase the number of high redshift clusters at , five additional clusters were taken from the EMSS (Gioia, Luppino, 1994), WARPS (Perlman et al., 2002; Horner et al., 2008), MACS (Ebeling et al., 2001) and NEP (Henry et al., 2006) surveys, which are compiled in the MCXC catalogue (Piffaretti et al., 2011).
The accuracy of our photometric redshift estimation is at and at . We see from Fig. 4 that our photometric redshift estimates, and therefore cluster IR luminosities, are not reliable at . Therefore, to identify Planck SZ sources, we use below only clusters with photometric redshift estimates .
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Burenin et al. (2007)
- 2[2] 1725612007
- 3Burenin (2015)
- 4[4] 411672015
- 5Burenin et al. (2016)
- 6[6] 422952016
- 7Bruzual, Charlot (2003)
- 8[8] 34410002003
