Issue 
A&A
Volume 617, September 2018



Article Number  A48  
Number of page(s)  17  
Section  Cosmology (including clusters of galaxies)  
DOI  https://doi.org/10.1051/00046361/201731489  
Published online  24 September 2018 
Planck intermediate results
LIII. Detection of velocity dispersion from the kinetic SunyaevZeldovich effect
^{1}
APC, AstroParticule et Cosmologie, Université Paris Diderot, CNRS/IN2P3, CEA/lrfu, Observatoire de Paris, Sorbonne Paris Cité, 10 rue Alice Domon et Léonie Duquet, 75205 Paris Cedex 13, France
^{2}
African Institute for Mathematical Sciences, 68 Melrose Road, Muizenberg, Cape Town, South Africa
^{3}
Agenzia Spaziale Italiana, Via del Politecnico snc, 00133 Roma, Italy
^{4}
Astrophysics Group, Cavendish Laboratory, University of Cambridge, J J Thomson Avenue, Cambridge, CB3 0HE, UK
^{5}
Astrophysics & Cosmology Research Unit, School of Mathematics, Statistics & Computer Science, University of KwaZuluNatal, Westville Campus, Private Bag X54001, Durban, 4000, South Africa
^{6}
CITA, University of Toronto, 60 St. George St., Toronto, ON M5S 3H8, Canada
^{7}
CNRS, IRAP, 9 Av. colonel Roche, BP 44346, 31028 Toulouse Cedex 4, France
^{8}
California Institute of Technology, Pasadena, California, USA
^{9}
Computational Cosmology Center, Lawrence Berkeley National Laboratory, Berkeley, California, USA
^{10}
DTU Space, National Space Institute, Technical University of Denmark, Elektrovej 327, DK2800 Kgs., Lyngby, Denmark
^{11}
Département de Physique Théorique, Université de Genève, 24, Quai E. Ansermet, 1211 Genève 4, Switzerland
^{12}
Departamento de Astrofísica, Universidad de La Laguna (ULL), 38206 La Laguna, Tenerife, Spain
^{13}
Departamento de Física, Universidad de Oviedo, C/Federico García Lorca 18, Oviedo, Spain
^{14}
Department of Astrophysics/IMAPP, Radboud University, PO Box 9010, 6500 GL Nijmegen, The Netherlands
^{15}
Department of Mathematics, University of Stellenbosch, Stellenbosch, 7602, South Africa
^{16}
Department of Physics & Astronomy, University of British Columbia, 6224 Agricultural Road, Vancouver, British Columbia, Canada
^{17}
Department of Physics & Astronomy, University of the Western Cape, Cape Town, 7535, South Africa
^{18}
Department of Physics and Astronomy, University of Sussex, Brighton, BN1 9QH, UK
^{19}
Department of Physics, Gustaf Hällströmin katu 2a, University of Helsinki, Helsinki, Finland
^{20}
Department of Physics, Princeton University, Princeton, New Jersey, USA
^{21}
Department of Physics, University of California, Santa Barbara, CA, USA
^{22}
Department of Physics, University of Illinois at UrbanaChampaign, 1110 West Green Street, Urbana, IL, USA
^{23}
Dipartimento di Fisica e Astronomia G. Galilei, Università degli Studi di Padova, via Marzolo 8, 35131 Padova, Italy
^{24}
Dipartimento di Fisica e Scienze della Terra, Università di Ferrara, Via Saragat 1, 44122 Ferrara, Italy
^{25}
Dipartimento di Fisica, Università La Sapienza, P. le A. Moro 2, Roma, Italy
^{26}
Dipartimento di Fisica, Università degli Studi di Milano, Via Celoria, 16, Milano, Italy
^{27}
Dipartimento di Fisica, Università degli Studi di Trieste, via A. Valerio 2, Trieste, Italy
^{28}
Dipartimento di Fisica, Università di Roma Tor Vergata, Via della Ricerca Scientifica 1, Roma, Italy
^{29}
European Space Agency, ESAC, Planck Science Office, Camino bajo del Castillo, s/n, Urbanización Villafranca del Castillo, Villanueva de la Cañada, Madrid, Spain
^{30}
European Space Agency, ESTEC, Keplerlaan 1, 2201 AZ Noordwijk, The Netherlands
^{31}
Gran Sasso Science Institute, INFN, viale F. Crispi 7, 67100 L’Aquila, Italy
^{32}
HGSFP and University of Heidelberg, Theoretical Physics Department, Philosophenweg 16, 69120 Heidelberg, Germany
^{33}
Haverford College Astronomy Department, 370 Lancaster Avenue, Haverford, Pennsylvania, USA
^{34}
Helsinki Institute of Physics, Gustaf Hällströmin katu 2, University of Helsinki, Helsinki, Finland
^{35}
INAF – OAS Bologna, Istituto Nazionale di Astrofisica  Osservatorio di Astrofisica e Scienza dello Spazio di Bologna, Area della Ricerca del CNR, Via Gobetti 101, 40129 Bologna, Italy
^{36}
INAF – Osservatorio Astronomico di Padova, Vicolo dell’Osservatorio 5, Padova, Italy
^{37}
INAF – Osservatorio Astronomico di Trieste, Via G.B. Tiepolo 11, Trieste, Italy
^{38}
INAF, Istituto di Radioastronomia, Via Piero Gobetti 101, 40129 Bologna, Italy
^{39}
INAF/IASF Milano, Via E. Bassini 15, Milano, Italy
^{40}
INFN – CNAF, viale Berti Pichat 6/2, 40127 Bologna, Italy
^{41}
INFN, Sezione di Bologna, viale Berti Pichat 6/2, 40127 Bologna, Italy
^{42}
INFN, Sezione di Ferrara, Via Saragat 1, 44122 Ferrara, Italy
^{43}
INFN, Sezione di Milano, Via Celoria 16, Milano, Italy
^{44}
INFN, Sezione di Roma 1, Università di Roma Sapienza, Piazzale Aldo Moro 2, 00185 Roma, Italy
^{45}
INFN, Sezione di Roma 2, Università di Roma Tor Vergata, Via della Ricerca Scientifica 1, Roma, Italy
^{46}
INFN/National Institute for Nuclear Physics, Via Valerio 2, 34127 Trieste, Italy
^{47}
Imperial College London, Astrophysics group, Blackett Laboratory, Prince Consort Road, London, SW7 2AZ, UK
^{48}
Institut d’Astrophysique Spatiale, CNRS, Univ. ParisSud, Université ParisSaclay, Bât. 121, 91405 Orsay Cedex, France
^{49}
Institut d’Astrophysique de Paris, CNRS (UMR7095), 98 bis Boulevard Arago, 75014 Paris, France
^{50}
Institute Lorentz, Leiden University, PO Box 9506, Leiden 2300 RA, The Netherlands
^{51}
Institute of Theoretical Astrophysics, University of Oslo, Blindern, Oslo, Norway
^{52}
Instituto de Astrofísica de Canarias, C/Vía Láctea s/n, La Laguna, Tenerife, Spain
^{53}
Instituto de Física de Cantabria (CSICUniversidad de Cantabria), Avda. de los Castros s/n, Santander, Spain
^{54}
Istituto Nazionale di Fisica Nucleare, Sezione di Padova, via Marzolo 8, 35131 Padova, Italy
^{55}
Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, California, USA
^{56}
Jodrell Bank Centre for Astrophysics, Alan Turing Building, School of Physics and Astronomy, The University of Manchester, Oxford Road, Manchester, M13 9PL, UK
^{57}
Kavli Institute for Cosmological Physics, University of Chicago, Chicago, IL 60637, USA
^{58}
Kavli Institute for Cosmology Cambridge, Madingley Road, Cambridge, CB3 0HA, UK
^{59}
LAL, Université ParisSud, CNRS/IN2P3, Orsay, France
^{60}
LERMA, CNRS, Observatoire de Paris, 61 Avenue de l’Observatoire, Paris, France
^{61}
Laboratoire de Physique Subatomique et Cosmologie, Université GrenobleAlpes, CNRS/IN2P3, 53, rue des Martyrs, 38026 Grenoble Cedex, France
^{62}
Laboratoire de Physique Théorique, Université ParisSud 11 & CNRS, Bâtiment 210, 91405 Orsay, France
^{63}
Lawrence Berkeley National Laboratory, Berkeley, CA, USA
^{64}
Low Temperature Laboratory, Department of Applied Physics, Aalto University, Espoo, 00076 AALTO, Finland
^{65}
MaxPlanckInstitut für Astrophysik, KarlSchwarzschildStr. 1, 85741 Garching, Germany
^{66}
Mullard Space Science Laboratory, University College London, Surrey, H5 6NT, UK
^{67}
NAOCUKZN Computational Astrophysics Centre (NUCAC), University of KwaZuluNatal, Durban, 4000 South Africa
email: ma@ukzn.ac.za
^{68}
Nicolaus Copernicus Astronomical Center, Polish Academy of Sciences, Bartycka 18, 00716 Warsaw, Poland
^{69}
Nordita (Nordic Institute for Theoretical Physics), Roslagstullsbacken 23, 106 91 Stockholm, Sweden
^{70}
SISSA, Astrophysics Sector, via Bonomea 265, 34136 Trieste, Italy
^{71}
San Diego Supercomputer Center, University of California, San Diego, 9500 Gilman Drive, La Jolla, CA, 92093, USA
^{72}
School of Chemistry and Physics, University of KwaZuluNatal, Westville Campus, Private Bag X54001, Durban, 4000, South Africa
^{73}
School of Physics and Astronomy, Cardiff University, Queens Buildings, The Parade, Cardiff, CF24 3AA, UK
^{74}
School of Physics and Astronomy, Sun YatSen University, 135 Xingang Xi Road, Guangzhou, PR China
^{75}
School of Physics, Indian Institute of Science Education and Research Thiruvananthapuram, Maruthamala PO, Vithura, Thiruvananthapuram, 695551 Kerala, India
^{76}
Simon Fraser University, Department of Physics, 8888 University Drive, Burnaby BC, Canada
^{77}
Sorbonne UniversitéUPMC, UMR7095, Institut d’Astrophysique de Paris, 98 bis Boulevard Arago, 75014 Paris, France
^{78}
Space Research Institute (IKI), Russian Academy of Sciences, Profsoyuznaya Str, 84/32, Moscow, 117997, Russia
^{79}
Space Science Data Center  Agenzia Spaziale Italiana, Via del Politecnico snc, 00133 Roma, Italy
^{80}
Space Sciences Laboratory, University of California, Berkeley, California, USA
^{81}
The Oskar Klein Centre for Cosmoparticle Physics, Department of Physics, Stockholm University, AlbaNova, 106 91 Stockholm, Sweden
^{82}
UPMC Univ. Paris 06, UMR7095, 98 bis Boulevard Arago, 75014 Paris, France
^{83}
Université de Toulouse, UPSOMP, IRAP, 31028 Toulouse Cedex 4, France
^{84}
Warsaw University Observatory, Aleje Ujazdowskie 4, 00478 Warszawa, Poland
Received:
2
July
2017
Accepted:
12
June
2018
Using the Planck fullmission data, we present a detection of the temperature (and therefore velocity) dispersion due to the kinetic SunyaevZeldovich (kSZ) effect from clusters of galaxies. To suppress the primary CMB and instrumental noise we derive a matched filter and then convolve it with the Planck foregroundcleaned “2DILC” maps. By using the Meta Catalogue of Xray detected Clusters of galaxies (MCXC), we determine the normalized rms dispersion of the temperature fluctuations at the positions of clusters, finding that this shows excess variance compared with the noise expectation. We then build an unbiased statistical estimator of the signal, determining that the normalized mean temperature dispersion of 1526 clusters is 〈(ΔT/T)^{2} 〉 = (1.64 ± 0.48) × 10^{−11}. However, comparison with analytic calculations and simulations suggest that around 0.7 σ of this result is due to cluster lensing rather than the kSZ effect. By correcting this, the temperature dispersion is measured to be 〈(ΔT/T)^{2}〉 = (1.35 ± 0.48) × 10^{−11}, which gives a detection at the 2.8 σ level. We further convert uniformweight temperature dispersion into a measurement of the lineofsight velocity dispersion, by using estimates of the optical depth of each cluster (which introduces additional uncertainty into the estimate). We find that the velocity dispersion is 〈υ^{2}〉 = (123 000 ± 71 000) (km s^{−1})^{2}, which is consistent with findings from other largescale structure studies, and provides direct evidence of statistical homogeneity on scales of 600 h^{−1} Mpc. Our study shows the promise of using crosscorrelations of the kSZ effect with largescale structure in order to constrain the growth of structure.
Key words: cosmic background radiation / largescale structure of Universe / galaxies: clusters: general / methods: data analysis
© ESO 2018
1. Introduction
The kinetic SunyaevZeldovich (hereafter kSZ; Sunyaev & Zeldovich 1972, 1980) effect describes the temperature anisotropy of the cosmic microwave background (CMB) radiation due to inverse Compton scattering of CMB photons off a moving cloud of electrons. The effect can be written as(1)
where σ_{T} is the Thomson crosssection, n_{e} is the electron density, υ · r̂ is the velocity along the lineofsight, and dl is the path length in the radial direction. By adopting a socalled “pairwise momentum estimator,” i.e., using the weights that quantify the difference in temperature between pairs of galaxies, the effect was first detected by Hand et al. (2012) using CMB maps from the Atacama Cosmology Telescope (ACT). The detection of the kSZ effect has been further solidified using the same pairwise momentum estimator with other CMB data, including Wilkinson Microwave Anisotropy Probe (WMAP) 9year Wband data, and Planck’s four foregroundcleaned maps (Planck Collaboration Int. XXXVII 2016), and again more recently using a Fourier space analysis (Sugiyama et al. 2018). These measurements represent detections at the 2–3 σ level. In addition, De Bernardis et al. (2017) applied the pairwise momentum estimator to the ACT data and 50 000 bright galaxies from BOSS DR11 catalogue, and obtained 3.6–4.1 σ C.L. detection. By using the same estimator to the South Pole Telescope (SPT) data and Dark Energy Survey (DES) data, Soergel et al. (2016) achieved the detection of kSZ signal 4.2 σ C.L. More recently Li et al. (2018) detected the pairwise kSZ signal for BOSS DR13 low mass group (M_{h} ≲ 10^{12}h^{−1}M_{⊙}) by using the 2DILC map (see Sect. 2.1.2). Besides the pairwise momentum estimator, in Planck Collaboration Int. XXXVII (2016) the kSZ temperature map (δT) was estimated from Planck fullmission data and crosscorrelated with the reconstructed linear velocity field data (υ · r̂) from the Sloan Digital Sky Survey (SDSSDR7) to compute the correlation function 〈 ΔT (υ · r̂)〉. For this crosscorrelation, 3.0–3.2 σ detections were found for the foregroundcleaned Planck maps, and 3.8 σ for the Planck 217GHz map. Following Planck Collaboration Int. XXXVII (2016), Schaan et al. (2016) detected the aggregated signal of the kSZ effect at 3.3 σ C.L. by crosscorrelating the velocity field from BOSS samples with the kSZ map produced from ACT observations. More recently, Hill et al. (2016) crosscorrelated the squared kSZ fields from WMAP and Planck, and the projected galaxy overdensity from the Widefield Infrared Survey Explorer (WISE), which led to a 3.8 σ detection. With advanced ACTPol and a future StageIV CMB experiment, the signaltonoise ratio (S/N) of the kSZ squared field and projected density field could reach 120 and 150, respectively (Ferraro et al. 2016), although these authors also cautioned that the results should be corrected for a bias due to lensing. These previous attempts to make various kinds of kSZ measurement are summarized in Table 1.
Recent measurements of the kinetic SunyaevZeldovich effect with crosscorrelations of various tracers of largescale structure.
There has been a lot of previous work investigating how to use kSZ measurements to determine the peculiar velocity field. This idea was first proposed by Haehnelt & Tegmark (1996), suggesting that on small angular scales the peculiar velocities of clusters could be inferred from CMB observations. Aghanim et al. (2001) estimated the potential uncertainty of the kSZ measurements due to contamination by the primary CMB and thermal SunyaevZeldovich (hereafter tSZ) effect. In Holzapfel et al. (1997), the peculiar velocities of two distant galaxy clusters, namely Abell 2163 (z = 0.201) and Abell 1689 (z = 0.181), were estimated through millimetrewavelength observations (the SZ Infrared Experiment, SuZIE). Furthermore, Benson et al. (2003) estimated the bulk flow using six galaxy clusters at z > 0.2 from the SuZIE II experiment in three frequency bands between 150 and 350 GHz, constraining the bulk flow to be <1410 km s^{−1} at 95% CL. In addition, Kashlinsky & AtrioBarandela (2000) and Kashlinsky et al. (2008, 2009) estimated peculiar velocities on large scales and claimed a “dark flow” (>≳1000 km s^{−1}) on Gpc scales. However, by combining galaxy cluster catalogues with Planck nominal mission foregroundcleaned maps, Planck Collaboration Int. XIII (2014) constrained the cluster velocity monopole to be 72 ± 60 km s^{−1} and the dipole (bulk flow) to be <254 km s^{−1} (95% CL) in the CMB rest frame. This indicates that the Universe is largely homogeneous on Gpc scales, consistent with the standard Λ cold dark matter (ΛCDM) scenario with adiabatic initial conditions^{1}.
This work represents the third contribution of the Planck^{2} Collaboration to the study of the kSZ effect. In Planck Collaboration Int. XIII (2014) we focused on constraining the monopole and dipole of the peculiar velocity field, which gives constraints on the largescale inhomogeneity of the Universe. In the second paper, Planck Collaboration Int. XXXVII (2016), we calculated the pairwise momentum of the kSZ effect and crosscorrelated this with the reconstructed peculiar velocity field 〈 ΔT (υ · r̂)〉, obtaining direct evidence of unbound gas outside the virial radii of the clusters. A followup paper modelled these results to reconstruct the baryon fraction and suggested that this unbound gas corresponds to all baryons surrounding the galaxies (HernándezMonteagudo et al. 2015). Even though the largescale bulk flow and monopole flow were not detected in Planck Collaboration Int. XIII (2014), the smallscale velocity dispersion in the nearby Universe, determined by the local gravitational potential field, might still be measurable. This is because the velocity of each galaxy comprises two components, namely the bulk flow components, which reflect the largescale perturbations, and a smallscale velocity dispersion component, which reflects perturbations due to the local gravitational potential (see, e.g., Watkins et al. 2009; Feldman et al. 2010; Ma & Scott 2013, 2014). Therefore, although the bulk flow of the galaxy clusters is constrained to be less than 254 km s^{−1}, the total velocity dispersion can still be large enough to be detected. With that motivation, in this paper we will look at a different aspect than in the previous two papers, namely focusing on 1point statistics of Planck data to constrain the temperature and velocity dispersion due to the kSZ effect. This topic is relevant for largescale structure, since the velocity dispersion that we are trying to measure can be used as a sensitive test for galaxy formation models (Ostriker 1980; Davies et al. 1983; Kormendy & Bender 1996; Kormendy 2001; MacMillan et al. 2006) and moreover, a numerical value for the smallscale dispersion often has to be assumed in studies of largescale flows (e.g., Ma et al. 2011; Turnbull et al. 2012). Providing such a statistical test through Planck’s fullmission foregroundcleaned maps is the main aim of the present paper.
This paper is organized as follows. In Sect. 2, we describe the Planck CMB data and the Xray catalogue of detected clusters of galaxies. In Sect. 3, we discuss the filter that we develop to convolve the observational map, and the statistical methodology that we use for searching for the kSZ temperaturedispersion signal. Then we present the results of our search along with relevant statistical tests. In Sect. 4, we discuss the astrophysical implications of our result, the conclusions being presented in the last section. Throughout this work, we adopt a spatially flat, CDM cosmology model, with the bestfit cosmological parameters given by Planck Collaboration XIII (2016): Ω_{m} = 0.309; Ω_{Λ} = 0.691; n_{s} = 0.9608; σ_{8} = 0.809; and h = 0.68, where the Hubble constant is H_{0} = 100 h km s^{−1} Mpc^{−1}.
2. Data description
2.1. Planck maps
2.1.1. Maps from the Planck Legacy Archive
In this work we use the publicly released Planck 2015 data^{3}. The kSZ effect gives rise to frequencyindependent temperature fluctuations that are a source of secondary anisotropies. The kSZ effect should therefore be present in all CMB foregroundcleaned products. Here we investigate the four Planck 2015 foregroundcleaned maps, namely the Commander, NILC, SEVEM, and SMICA maps. These are the outputs of four different componentseparation algorithms (Planck Collaboration Int. XXXVII 2016) and have a resolution of θ_{FWHM} = 5 arcmin. SMICA uses a spectralmatching approach, SEVEM adopts a templatefitting method to minimize the foregrounds, NILC is the result of an internal linear combination approach, and Commander uses a parametric, pixelbased Monte Carlo Markov chain technique to project out foregrounds (Planck Collaboration XII 2014; Planck Collaboration IX 2016; Planck Collaboration X 2015). All of these maps are produced with the intention of minimizing the foreground contribution, but there could nevertheless be some residual contamination from the tSZ effect, as well as other foregrounds (e.g., the Galactic kSZ effect, see Waelkens et al. 2008). We use the HEALPix package (Górski et al. 2005) to visualize and mainpulate the maps.
2.1.2. The 2DILC map
The 2DILCPlanck CMB map has the additional benefit of being constructed to remove contamination from the tSZ effect, provided that the tSZ spectral energy distribution is perfectly known across the frequency channels. The 2DILC CMB map has been produced by taking the Planck 2015 data and implementing the “constrained ILC” method developed in Remazeilles et al. (2011a). This componentseparation approach was specifically designed to cancel out in the CMB map any residual of the tSZ effect towards galaxy clusters by using spectral filtering, as we now describe.
For a given frequency band i, the Planck observation map x_{i} can be modelled as the combination of different emission components:(2)
where s_{CMB}(r̂) is the CMB temperature anisotropy at pixel r̂, s_{tSZ}(r̂) is the tSZ fluctuation in the same direction, and n_{i}(p) is a “nuisance” term including instrumental noise and Galactic foregrounds at frequency i. The CMB fluctuations scale with frequency through a known emission law parameterized by the vector a, with nine components, accounting for the nine Planck frequency bands. The emission law of the tSZ fluctuations is also known and can be parameterized by the scaling vector b in the Planck frequency bands. The kSZ signal is implicitly included in the CMB fluctuations, since CMB anisotropies and kSZ fluctuations share the same spectral signature.
Similar to the standard NILC method (Basak & Delabrouille 2012, 2013), the 2DILC approach makes a minimumvarianceweighted linear combination of the Planck frequency maps. Specifically, under the condition that the scalar product of the weight vector w and the CMB scaling vector a is equal to unity, i.e., , which guarantees the conservation of CMB anisotropies in the filtering. However, 2DILC (Remazeilles et al. 2011a) generalizes the standard NILC method by offering an additional constraint for the ILC weights to be orthogonal to the tSZ emission law b, while guaranteeing the conservation of the CMB component. The 2DILC CMB estimate is thus given by(3)
such that the variance of Eq. (3) is minimized, with(4) (5)
Benefiting from the knowledge of the CMB and tSZ spectral signatures, the weights of the 2DILC are constructed in order to simultaneously yield unit response to the CMB emission law a (Eq. (4)) and zero response to the tSZ emission law b (Eq. (5)). The residual contamination from Galactic foregrounds and instrumental noise is controlled through the condition (Eq. (3)). The exact expression for the 2DILC weights was derived in Remazeilles et al. (2011a) by solving the minimization problem (Eqs. (3)–(5)):(6)
where are the coefficients of the frequencyfrequency covariance matrix of the Planck channel maps; in practice we compute this locally in each pixel p as(7)
Here the pixel domain D(p) (referred to as “super pixels”) around the pixel p is determined by using the following procedure: the product of frequency maps x_{i} and x_{j} is convolved with a Gaussian kernel in pixel space in order to avoid sharp edges at the boundaries of super pixels that would create spurious power (Basak & Delabrouille 2012, 2013).
Before applying the 2DILC filter (Eq. (6)) to the Planck 2015 data, we first preprocess the data by performing pointsource “inpainting” and wavelet decomposition, in order to optimize the foreground cleaning. In each Planck channel map we mask the pointsources detected at S/N > 5 in the Second Planck Catalogue of Compact Sources PCCS2 (Planck Collaboration XXVI 2016). The masked pixels are then filled in by interpolation with neighbouring pixels through a minimum curvature spline surface inpainting technique, as implemented in Remazeilles et al. (2015). This preprocessing of the pointsource regions will guarantee reduction of the contamination from compact foregrounds in the kSZ measurement.
The inpainted Planck maps are then decomposed into a particular family of spherical wavelets called “needlets” (see, e.g., Narcowich et al. 2006; Guilloux et al. 2009). The needlet transform of the Planck maps is performed as follows. The spherical harmonic coefficients a_{i, ℓm} of the Planck channel maps x_{i} are bandpass filtered in multipole space in order to isolate the different ranges of angular scales in the data. The 2DILC weights (Eq. (6)) are then computed in pixel space from the inverse spherical harmonic transform of the bandpassfiltered a_{i, ℓm} coefficients. The frequencyfrequency covariance matrix in Eq. (7) is actually computed on the bandpassfiltered maps. In this way, component separation is performed for each needlet scale (i.e., range of multipoles) independently. Due to their localization properties, the needlets allow for a filtering in both pixel space and multipole space, therefore adapting the componentseparation procedure to the local conditions of contamination in both spaces (see Delabrouille et al. 2009; Remazeilles et al. 2011b; Basak & Delabrouille 2012, 2013).
The top left panel of Fig. 1 shows the result of stacking 3° × 3° patches of the NILCPlanck CMB map in the direction of known galaxy clusters, while the top right panel shows the result of stacking the 2DILCPlanck CMB map in the direction of the same set of galaxy clusters^{4}. The Planck SZ sample provides a very stringent test, because these are the places on the sky where Planck detected a significant y signature. We see that stacking of the NILC CMB map shows a significant tSZ residual effect in the direction of galaxy clusters. Conversely, the stacking of the 2DILCPlanck CMB map (right panel of Fig. 1) appears to show substantially reduced tSZ residuals, due to the 2DILC filtering. In the bottom panels of Fig. 1 we show the results of the stacking procedure for the specific cluster catalogue that we will be using for the main analysis in this paper (see next section for details).
Fig. 1.
Top left: stack of the NILC CMB map in the directions of Planck SZ (PSZ) galaxy clusters. Top right: stack of the 2DILC CMB map in the direction of PSZ galaxy clusters. This sample provides a very stringent test of the tSZ leakage, since the PSZ positions are the known places on the sky with detectable SZ signal. The stacked NILC CMB map clearly shows an excess in the centre, which is due to residual contamination from the tSZ effect, while the 2DILC CMB map has a signature in the centre that is consistent with the strength of other features in the stacked image. Bottom left: stack of the NILC CMB map in the directions of MCXC clusters. Bottom right: stack of the 2DILC CMB map in the direction of 1526 MCXC clusters (see Sect. 2.2 for the detail of the catalogue). For a different set of sky positions, the results are broadly consistent with those for the PSZ clusters. All these maps are 3° × 3° in size, and use the same colour scale. 
The profiles of the stacked patches are plotted in Fig. 2. The excess of power due to tSZ residuals in the NILC CMB map would clearly lead to a significant bias in any attempt to detect the kSZ signal at the positions of the galaxy clusters. As a baseline reference, the flux profile of the Planck 217GHz map, stacked in the directions of these galaxy clusters, is also plotted in Fig. 2 (green squares). The tSZ signal should in principle vanish in observations at 217 GHz, since that is effectively the null frequency for the tSZ signature; in practice it is nonzero in the Planck 217GHz map because of the broad spectral bandpass. In fact there is an offset of about 20 μ K in the flux profile of the stacked 217GHz map at the position of PSZ clusters. There is also still a residual offset in the 2DILC CMB map; however, it is smaller by a factor of about 2 than the tSZ signal in the baseline Planck 217GHz map (see top panel of Fig. 2), and dramatically better than for the NILC map. This suggests that the method employed for the 2DILC map was successful in removing the tSZ signal.
Fig. 2.
Profiles for stacked patches (see Fig. 1) of the NILC CMB map (black diamonds) and the 2DILC CMB map (blue triangles) at the positions of PSZ clusters (top panel) and MCXC clusters (bottom panel). The profile of the stacked Planck 217GHz map is also shown as a reference (green squares). The central deficit in the flux profile of the stacked NILC CMB map (black diamonds) is due to residual tSZ contamination. 
The residual flux of the stacked 2DILC CMB map in the direction of galaxy clusters can be interpreted as the result of possibly imperfect assumptions in the 2DILC filter and the exact tSZ spectral shape across the Planck frequency bands. There may be several reasons behind incomplete knowledge of the tSZ spectrum: detector bandpass mismatch; calibration uncertainties; and also relativistic tSZ corrections. In addition, even if the kSZ flux is expected to vanish on average when stacking inward and outwardmoving clusters in a homogeneous universe, there is still a potential selection bias (since we use a selected subset of clusters for stacking) that may result in a nonzero average kSZ residual in the offset of the 2DILC map. Although it is not easy to estimate the size of all these effects, we are confident that they cannot be too large because the residual offset in the 2DILC map is negligible compared to the tSZ residuals in Planck CMB maps, and smaller by a factor of 2 with respect to the baseline Planck 217GHz map.
Regarding residual Galactic foreground contamination, we checked that the angular power spectrum of the 2DILC CMB map on the 60% of the sky that is unmasked is consistent with the angular power spectrum of the PlanckSMICA CMB maps. There is therefore no obvious excess of power due to Galactic emission. We also checked the amount of residual dust contamination of the kSZ signal on small angular scales in the direction of the galaxy clusters, where dusty starforming galaxies are present (Planck Collaboration Int. XLIII 2016). Considering the Planck 857GHz map as a dust template, we scaled it across the Planck frequency bands using a modified blackbody spectrum with bestfit values from Planck Collaboration Int. XLIII (2016), i.e., β = 1.5 and T = 24.2 K. This provides dust maps at each frequency band. We then applied the ILC weights that go into the 2DILC CMB+kSZ map (Eq. (6)) to the thermal dust maps. This provides an estimate of the map of the residual dust contamination in the 2DILC map. We then stacked the residual dust map in the direction of the galaxy clusters from either the PSZ or the MCXC catalogue, and computed the profile of the stacked patch as in Fig. 2. We found that the residual flux from the dust stacked in the direction of the galaxy clusters is compatible with zero.
Residual cosmic infrared background (hereafter CIB) and instrumental noise in the CMB maps will add some scatter to the measured kSZ signals in the directions of galaxy clusters, but should not lead to any bias in the stacked profile. However, any additional source of extra noise will lead to bias in the variance of the stacked profile. Since CIB and noise are not spatially localized on the sky (unlike kSZ and tSZ signals) this bias can be estimated using offcluster positions, e.g., for the matchedfiltering analysis performed in Sect. 3.2.
In order to quantify the amount of residual noise in the 2DILC CMB map, we apply the 2DILC weights (calculated from the Planck fullsurvey maps) to the first and second halves of each stable pointing period (also called “rings”). In the halfdifference of the resulting “first” and “second” 2DILC maps, the sky emission cancels out, therefore leaving an estimate of the noise contamination in the 2DILC CMB maps, constructed from the fullsurvey data set.
The 2DILC CMB map shows approximately 10% more noise than the NILC CMB map; this arises from the additional constraint imposed in the 2DILC of cancelling out the tSZ emission. At the cost of having a slightly higher noise level, the 2DILC CMB map benefits from the absence of bias due to tSZ in the directions of galaxy clusters. For this reason, the 2DILC CMB map is particularly well suited for the extraction of the kSZ signal in the direction of galaxy clusters and we shall focus on it for the main results of this paper.
2.2. The MCXC Xray catalogue
To trace the underlying baryon distribution, we use the Meta Catalogue of Xray detected Clusters of galaxies (MCXC), which is an allsky compilation of 1743 allsky ROSAT surveybased samples (BCS, Ebeling et al. 1998, 2000; CIZA, Ebeling et al. 2010; Kocevski et al. 2007; MACS, Ebeling et al. 2007; NEP, Henry et al. 2006; NORAS, Böhringer et al. 2000; REFLEX, Böhringer et al. 2004; SCP, Cruddace et al. 2002) along with a few other catalogues (160SD, Mullis et al. 2003; 400SD, Burenin et al. 2007; EMSS, Gioia & Luppino 1994; Henry 2004; SHARC, Romer et al. 2000; Burke et al. 2003; WARPS, Perlman et al. 2002; Horner et al. 2008). We show stacks and profiles for this catalogue on the Planck map in Figs. 1 and 2. While selecting sources from this catalogue, we use the luminosity within R_{500} (the radius of the cluster within which the density is 500 times the cosmic critical density), L_{500}, and restrict the samples to have 1.5 × 10^{33} W < L_{500} < 3.7 × 10^{38} W within the band 0.1–2.4 keV (see Piffaretti et al. 2011). As well as L_{500}, for each cluster the catalogue gives M_{500}, the mass enclosed within R_{500} at redshift z, i.e., , estimated using the empirical relation in Arnaud et al. (2010). Further details of catalogue homogenization and calibration are described in Piffaretti et al. (2011) and Planck Collaboration Int. XIII (2014).
For each cluster in the MCXC catalogue, the properties we use in the rest of this paper are the sky position (Galactic coordinates l, b), the redshift z, and the mass M_{500}. In Sect. 4 we will use M_{500} and z to estimate the optical depth for each cluster.
Since in the CMB map, the Galactic plane region is highly contaminated by foreground emission, we use the Planck Galactic and pointsource mask to remove 40% of the sky area. The number of MCXC sources outside the sky mask is N_{c} = 1526 (which we use throughout the paper) and their spatial and redshift distributions are shown in Fig. 3. The fullsky distribution is presented in the left panel of Fig. 3, and one can see that the distribution of MCXC clusters is roughly uniform outside the Galactic mask. The redshift of MCXC clusters peaks at z = 0.09, with a long tail towards higher redshift, z >≳ 0.4.
Fig. 3.
Left: fullsky distribution of 1526 MCXC Xray clusters (Piffaretti et al. 2011; Planck Collaboration Int. XIII 2014) in Galactic coordinates. The dark blue area is the masked region, and the clusters are shown in orange. Right: redshift histogram of 1526 Xray clusters, with bin width Δz = 0.025. 
3. Methodology and statistical tests
3.1. Matchedfilter technique
The foregroundcleaned CMB maps (SEVEM, SMICA, NILC, Commander, and 2DILC) contain mainly the primary CMB and kSZ signals, so in order to optimally characterize the kSZ signal, we need to use a spatial filter to convolve the maps in order xto downweight the CMB signal. Here we use the matchedfilter technique (e.g., Tegmark & de OliveiraCosta 1998; Ma et al. 2013), which is an easily implemented approach for suppressing the primary CMB and instrumental noise.
Most of Planck’s SZclusters are unresolved, so we treat them as point sources on the sky. In this limit, if cluster i has flux S_{i} at sky position r̂_{i}, the sky temperature ΔT(r̂) can be written as(8)
where δ_{D} is the Dirac delta function, c is the conversion factor between flux and temperature, and the spherical harmonics characterize the true CMB fluctuations. The sky signal, obtained from the Planck telescope, is(9)
where is the true CMB signal convolved with the beam plus the detector noise, i.e., (assuming that this is the only source of noise). The beam function of Planck foregroundcleaned maps in ℓspace is close to a Gaussian with θ_{FWHM} = 5 arcmin, i.e., , with . Residual foregrounds in the Planck CMB maps and in the 2DILC CMB map have been minimized in the componentseparation algorithms, as demonstrated in Planck Collaboration IX (2016) for the public Planck CMB maps and in Planck Collaboration Int. XIII (2014) for the 2DILC CMB map. Figure 4 compares the angular power spectrum, C_{ℓ}, directly estimated from the map by using the pseudoC_{ℓ} estimator (Hivon et al. 2002), and the spectrum predicted by using the bestfit CDM model and noise template. One can see that the measured spectral data scatter around the predicted spectrum, and that the two spectra are quite consistent with each other.
Fig. 4.
Measured (black dots) and predicted (red line) power spectra from the Planck2DILC map. The predicted spectrum is based on the bestfitting ΛCDM model convolved with the squared beam , with the noise added. These are estimated using the pseudoC_{ℓ} estimator described in Hivon et al. (2002). 
In order to maximize our sensitivity to SZ clusters, we further convolve ΔT^{ obs}(r̂) with an optimal filter W_{ℓ}:(10)
where we are seeking the form of W_{ℓ} that will maximize cluster S/N. In the direction of each cluster, the filtered signal is(11)
and we want to vary W_{ℓ} to minimize the ratio(12)
where , and we take to be the CDM model power spectrum. Since A in Eq. (11) is a constant, we minimize Eq. (12) by adding a Lagrange multiplier to the numerator (see, e.g., Ma et al. 2013), i.e., we minimize(13)
which we plot in Fig. 5 as a black line, along with the primary CMB C_{ℓ}, the noise map, and their sum. In the filtering process, we use the normalized filter , where . One can see that the filter function W_{ℓ} gives lower weight in the primary CMB domain while giving more weight in the cluster regime, ℓ >≳ 2000. We then convolve the five Planck foregroundcleaned maps with this W_{ℓ} filter, noting that the noise power spectrum N_{ℓ} in Eq. (14) of each foregroundcleaned map is estimated by using its corresponding noise map. After we perform this step, the primary CMB features are highly suppressed (Fig. 6), and the whole sky looks essentially like a noisy map, although it still contains the kSZ information of course.
Fig. 5.
Optimal matched filter (black line) for pointsource detection in the Planck2DILC map (Eq. (14)). For comparison, the power spectra of the CMB signal (red line) and noise map (blue dashed line) are shown, along with their sum (brown line). 
3.2. Statistical method and tests of robustness
We now proceed to estimate the kSZ temperature dispersion and perform various tests. The filtered map contains the kSZ signal and residual noise, and from this we plot the histogram of 1526 ΔT/T values at the cluster positions (see red bars in Fig. 7). We can also randomly select the same number of pixels on the sky and plot a histogram for that. The two histograms have almost zero mean value (Table 2), but in the real cluster positions yield a larger variance than the random selections, i.e., the real cluster positions give a slightly broader distribution than for the randomly selected positions (Table 2). We also show results for the skewness and kurtosis of the two samples in Table 2, and one can see that for these statistics the real cluster positions also give larger values than for the randomlyselected positions. This suggests that there may be additional tests that could be performed to distinguish the real cluster kSZ signals; however, we leave that for future studies, and for the rest of this paper we just focus on investigating whether the slight broadening of the distribution is due to the kSZ effect.
Statistics of the values of (ΔT/T) × 10^{5} at the true 1526 cluster positions and for 1526 randomlyselected positions.
Fig. 7.
Histograms of 1526 ΔT/T values of 2DILC map at the cluster catalogue positions (red bars), and randomly selected positions (black bars). The statistics of the true cluster positions and random positions can be found in Table 2. 
3.2.1. Test of thermal SunyaevZeldovich effect residuals
The first test we want to perform is to check whether the measured kSZ (ΔT/T) value at each cluster position suffers from residuals of the tSZ effect. The mapmaking procedures of SMICA, NILC, SEVEM, and Commander minimize the variance of all nonCMB contribution to the map, but they are not designed to null the tSZ component. By contrast, the 2DILC map is designed to also null the tSZ contribution, and therefore should provide a cleaner measurement of the kSZ effect (but with a slightly higher noise level).
We first choose 5000 randomly selected catalogues from each Planck foregroundcleaned map, each being a collection of 1526 random positions on the sky. We then calculate the average value of (ΔT/T) for each random catalogue and plot the resulting histograms in Fig. 8. The five different colours of (overlapping) histogram represent the different Planck maps. One can see that they are all centred on zero, with approximately the same widths. Since the 2DILC map has nulled the tSZ component in the map, it does not minimize the variance of all foreground components and as a result, its width in Fig. 8 (2.86 × 10^{−7}) is slightly larger than for all other maps; in these units, the 1 σ width of the histograms for SMICA, NILC, SEVEM and Commander are 2.49 × 10^{−7}, 2.48 × 10^{−7}, 2.53 × 10^{−7}, and 2.62 × 10^{−7}, respectively. This indicates that the noise level in the filtered 2DILC map is slightly higher than for the other four maps.
Fig. 8.
Histogram of the values of 5000 random catalogues on the sky (each having N_{c} = 1526), with different colours representing different Planck foregroundcleaned maps. The 68% width of the 2DILC, SMICA, NILC, SEVEM and Commander histograms are 2.86, 2.49, 2.48, 2.53, 2.62 (×10^{−7}), respectively. The vertical lines represent the average ΔT/T values at true MCXC cluster positions for each map. One can see that only for the 2DILC map is this value within the 68% range of the random catalogue distribution, while others are quite far off. This indicates that, except for the 2DILC map, the public Planck maps have residual tSZ contamination at the cluster positions. 
We then calculate the average value of ΔT/T at the true cluster positions for the five Planck foregroundcleaned maps as the vertical bars in Fig. 8. One can see that only the average value of the 2DILC map lies close to zero and within the 68% width of the noise histogram, while the values of all other maps are quite far from the centre of the noise distribution. This strongly suggests that at each of the true cluster positions the (ΔT/T) value contains some contribution from the tSZ effect, so that the tSZ effect contributes extra variance to the foregroundcleaned maps.
3.2.2. Test with random positions
We now want to test whether this slight broadening of the distribution is a statistically significant consequence of the kSZ effect. So for the 5000 randomly selected catalogues, we calculate the scatter of the 1526 (ΔT/T) values. We then plot (in Fig. 9) the histogram of 5000 rms values of these random catalogues, and mark the rms value of the true MCXC cluster positions for reference. One can see that the mean of the 5000 rms values of the random catalogues is 1.10 × 10^{−5}, and that the scatter of the 5000 rms values of the random catalogue has a width around 2.15 × 10^{−7}. The rms value of the 1526 true MCXC position is 1.17 × 10^{−5}, larger than the mean value at more than the 3 σ level.
Fig. 9.
Distribution of the rms for 5000 mock catalogues (yellow histogram); here each catalogue consists of N_{c} randomly chosen positions on the filtered 2DILC map. The mean and rms of the 5000 random catalogues are 1.10 × 10^{−5} (shown as the black dashed vertical line) and 2.15 × 10^{−7}, respectively. For the N_{c} true MCXC cluster positions, the rms is 1.17 × 10^{−5} (red vertical dashed line). 
In Table 3, we list the rms value for the true sky positions of the 1526 MCXC catalogue sources (σ_{MCXC}, the mean (σ_{ran}), and standard deviation (σ(σ_{ran})) for 5000 random catalogues for different foregroundcleaned maps. One can see that although the absolute value of each map varies somewhat, the second column (σ_{ran}) is consistently smaller than the first column (σ_{MCXC}) by roughly 0.07–0.13 × 10^{−5}, which, specifically for 2DILC is about 3 times the scatter of the rms among catalogues (σ(σ_{ran})). This consistency strongly suggests that the kSZ effect contributes to the extra dispersion in the convolved ΔT/T maps at the cluster positions (since the 2DILC map is constructed to remove the tSZ effect). If we use the SMICA, SEVEM, NILC, and Commander maps, the detailed values will vary slightly due to the different calibration schemes of the maps, but the detection remains consistently there. The difference between σ_{MCXC} and σ_{ran} is slightly larger in the SMICA, NILC, and SEVEM maps due to residual tSZ contamination, shown as the vertical bars in Fig. 8.
rms values for the true sky positions of 1526 MCXC catalogue clusters (σ_{MCXC}), along with the mean (σ_{ran}) and scatter (σ(σ_{ran})) of the values of the rms for 5000 random catalogues, where each catalogue consists of 1526 random positions on the sky.
This all points towards the broadening of the ΔT/T histogram being a consequence of the kSZ effect; hence we identify it as additional temperature dispersion arising from the scatter in cluster velocities detected through the kSZ effect. In Sect. 4, we will interpret this effect in terms of the lineofsight velocity dispersion of the SZ clusters, which is an extra variance predicted in linear perturbation theory in the standard picture of structure formation (Peebles 1980).
3.2.3. Test with finite cluster size
We now want to test how much our results depend on the assumption that SZ clusters are point sources. A cluster on the sky appears to have a radius of θ_{500}, which is equal to θ_{500} = R_{500}/D_{A}, where R_{500} is the radius from the centre of the cluster at which the density contrast is equal to 500 and D_{A} is the angular diameter distance to the cluster. The peak in the distribution of θ_{500} values for the MCXC clusters lies at around 3 arcmin, so we multiply the filter function (Eq. (14)) with an additional “cluster beam function” , where . We pick three different values for the cluster size, namely θ_{500} = 3, 5, and 7 arcmin, and see how our results change.
We list our findings in Table 4; one can see that the detailed values for three cases are slightly different from those of the pointsource assumptions, but the changes are not dramatic. More importantly, the offsets between σ_{MCXC} and σ_{ran} stay the same for various assumptions of cluster size. Therefore the detection of the temperature dispersion due to the kSZ effect does not strongly depend on the assumption of clusters being point sources.
3.3. Statistical results
3.3.1. Statistics with the uniform weight
We now want to perform a more quantitative calculation of the significance of detection. Since the convolved map mainly consists of the kSZ signal at the cluster positions plus residual noise, we write the observed temperature fluctuation at the cluster positions as(15)
where δ, s, and n represent the observed ΔT/T value, the kSZ signal contribution, and the residual noise, respectively, all of which are dimensionless quantities. Now we define the estimator as(16)
where the summation includes all of the N_{c} = 1526 cluster positions. For the first term δ_{i}, we use the N_{c} true cluster position as the measurement of each observed ΔT/T. For the second term, we randomly select N_{c} pixels outside the Galactic and pointsource mask that are not cluster positions. The calculation of the first term is fixed, whereas the second term depends on the N_{c} random positions we choose. Each randomly selected set of N_{c} positions corresponds to a mock catalogue, which leads to one value of s^{2}. We do this for 5000 such catalogues, where each mock catalogue has a different noise part (n̂_{i}) in Eq. (16), but the same observed δ_{i}. Then we plot the histogram of s^{2} values for these catalogues in the left panel of Fig. 10. One can see that the s^{2} distribution is close to a Gaussian distribution with mean and error being s^{2} = (1.64 ± 0.48) × 10^{−11}.
Fig. 10.
Left: distribution of 5000 values of s^{2} with uniform weight (Eq. (16)) on each position for the 2DILC map. Right: distribution of υ^{2} calculated from Eq. (30), after subtracting the lensing shift. The distribution has P(υ^{2} < 0) = 5.3%, corresponding to a (1sided) detection of 1.6 σ. We have tested with 50 000 values of s^{2} and the results are consistent with those from 5000 values. 
One can use a complementary method to obtain the mean and variance of , i.e., E[s^{2}] and V[s^{2}]. We lay out this calculation in Appendix A, where we directly derive these results:(17)
Here μ_{2} and μ_{4} are the second and fourth moments of the corresponding random variables.
For the moments of δ, we use the measurements at the 1526 cluster positions. For the estimate of the noise, we take all of the unmasked pixels of the convolved sky. In order to avoid selecting the real cluster positions, we remove all pixels inside a 10 arcmin aperture around each cluster. These “holes” at each cluster position constitute a negligible portion of the total unmasked pixels, and our results are not sensitive to the aperture size we choose. As a result, we have approximately 3 × 10^{7} unmasked pixels to sample the noise. We then substitute the values into Eq. (17) to obtain the expectation values and variances.
In Table 5, we list the mean and rms value of . Comparing with the 2DILC map, one can see that the SMICA, NILC, and SEVEM maps give larger values of E[s^{2}] and therefore apparently higher significance levels, which we believe could be due to the fact that the residual tSZ effect in these maps contributes to the signal. However, the Commander map gives a reasonable estimate of the dispersion, since it appears to be less contaminated by tSZ residuals (see Fig. 8). As discussed in Sect. 2.1.2, the mapmaking procedure of the 2DILC product enables us to null the tSZ effect so that the final map should be free of tSZ, but with larger noise. This is the reason that we obtain a somewhat lower significance in Table 5 for 2DILC compared to some of the other maps. We will therefore mainly quote this conservative detection in the subsequent analysis.
Statistics of the variables due to the kSZ effect for different CMB maps.
3.3.2. Statistics with different weights
The results so far have been found using the same weights for each cluster position. We now examine the stability of the detection using weighted stacking. In Eq. (16) we defined stacking with uniform weights, which can be generalized to(18)
where w_{i} is the weight function. We certainly expect “larger” clusters to contribute more to the signal, but it is not obvious what cluster property will be best to use. In Table 6, we try different weighting functions w_{i}, with the first row being the uniform weight, which is equivalent to Eq. (16). In addition, we try as different choices of weighting function the optical depth τ and its square τ^{2}^{5}, the angular size θ_{500} and its square , the luminosity L_{500} and its square , and the mass M_{500} and its square . Since some of these may give distributions of that deviate from Gaussians, we also calculate the frequencies for finding smaller than zero, . The smaller this Pvalue is, the more significant is the detection.
Statistics of the weighted variables for different choices of weights in the 2DILC map.
From Table 6, we see that most weighting choices are consistent with uniform weighting though with reduced significance of the detection, the exceptions being the choices of θ_{500} or . For w_{i} = θ_{500,i} and , we have Pvalues of 0.0002 and 0.0039, respectively, yielding (1sided) significance levels of 3.5 σ and 2.7 σ. Their distributions deviate slightly from a Gaussian, with a tail toward smaller values.
The increased detection of excess variance using θ_{500} weighting stems from our choice of using a single cluster beam function for all clusters. In Sect. 3.2.3 we tested the robustness of our results to the choice of cluster beam function, finding little dependence. Nevertheless, such a test assumed all clusters had the same angular size, while in reality there is a large spread in the angular sizes of the clusters. By weighting with θ_{500} we are able to recover some of this lost signal in a quick and simple way, which we tested by comparing results for larger clusters versus smaller clusters. Despite this, we find that the increased significance is mainly due to the increased value for and not a decrease in the noise. This tension may be evidence of systematic effects in the data or potentially some noise coming from residual tSZ signal in the 2DILC map; this should be further investigated when better data become available.
3.3.3. Statistics with split samples
We further split the samples into two groups, according to their median values of M_{500}, R_{500} and L_{500}. The distributions of these values are shown in Fig. 11. One can see that our samples span several orders of magnitude in mass, radius, and luminosity, but that the median values of the three quantities quite consistently split the samples into two. We recalculate the statistics as defined in Eq. (16) for the two subsamples and list our results in Table 7.
Fig. 11.
Left: R_{500} versus M_{500} for the sample. Right: L_{500} versus M_{500} for the sample. The vertical and horizontal lines show the median values of R_{500}, M_{500}, and L_{500}. 
Statistics of split samples. Here the median values of the samples are (M_{500})_{mid} = 1.68 × 10^{14}M_{⊙}, (L_{500})_{mid} = 2.3 × 10^{10}L_{⊙}, and (R_{500})_{mid} = 0.79 Mpc.
We see that the groups with lower values of L_{500}, M_{500}, and R_{500} give detections of the kSZ temperature dispersion at 1.1 σ, 1.2 σ, and 0.7 σ, whereas the higher value subsets give higher S/N detections. It is clear that the signal we see is dominated by the larger, more massive, and more luminous subset of clusters, as one might expect. In principle one could use this information (and that of the previous subsection) to devise an estimator that takes into account the variability of cluster properties in order to further maximize the kSZ signal; however, a cursory exploration found that, for our current data, improvements will not be dramatic. We therefore leave further investigations for future work.
3.4. Effect of lensing
An additional effect that could cause a correlation between our tSZfree CMB maps and clusters comes from gravitational lensing (as discussed in Ferraro et al. 2016). It is therefore important to determine what fraction of our putative kSZ signal might come from lensing, and we estimate this in the following way. The MCMX CMB temperature variance is set by the integrated local CMB power spectrum at these positions. Lensing by the clusters magnifies the CMB, locally shifting scales with respect to the global average, potentially introducing a lensing signal in the variance shift. If we were to compare the CMB variance between cluster lensed and unlensed skies unlimited by resolution, no difference would be seen; lensing is merely a remapping of points on the sky, and hence does not affect onepoint measures, with local shifts in scales in the power spectrum being compensated by subtle changes in amplitude, keeping the variance the same. However, as noted by Ferraro et al. (2016), the presence of finite beams and of the filtering breaks this invariance and the relevance of the lensing effects needs to be assessed. Crudely, the clusterlensing signal α · ∇T (with α the deflection due to the cluster) can be as large as 5 μK (Lewis & Challinor 2006), hence potentially contributing ≃ 25 μK^{2} to the observed variance , i.e., around the 1 σ level.
We obtain a more precise estimate with a CMB simulation as follows. After generating a Gaussian CMB sky with lensed CMB spectra, we lens the CMB at each cluster position according to the deflection field predicted for a standard halo profile (Navarro et al. 1996; Dodelson 2004) of the observed redshift and estimated mass. We use for this operation a bicubic spline interpolation scheme, on a highresolution grid of 0.4 arcmin (N_{side} = 8192), using the python lensing tools^{6}. We then add the 2DILC noisemap estimate to the CMB. We can finally compare the temperature variance at the cluster positions before and after cluster lensing. We show our results in Fig. 12. The background (yellow) histogram shows the value of the dispersion of the 5000 random catalogues on the simulated map with 2DILC noise added. The purple (red) line indicates the value of dispersion of the map without (with) cluster lensing added. One can see that with cluster lensing added the value of dispersion is shifted to a higher value by 1.4 × 10^{−7}, which is somewhat less than the rms of the random catalogue (2.1 × 10^{−7}). This indicates that the lensing causes a roughly 0.7 σ shift in the width of histogram. From Fig. 12, we can calculate the lensingshifted ŝ^{2} as(19)
Fig. 12.
Histogram of the values of 5000 random catalogues (each having N_{c} = 1526) on simulated lensed skies with added 2DILC noise. The mean and rms for the 5000 mock catalogues are 1.07 × 10^{−5} and 2.13 × 10^{−7}, respectively. The vertical dashed lines indicate the values at the MCXC cluster positions before and after the clusterlensing effect is added; the values are 1.047 × 10^{−5} before (purple line) and 1.061 × 10^{−5} after (red line). 
Therefore, using uniform weights, by subtracting the , the temperature dispersion in the 2DILC map is measured to be(20)
which is thus detected at the 2.8σ level.
3.5. Comparison with other kSZ studies
Table 1 summarizes all of the previous measurements of the kSZ effect coming from various crosscorrelation studies. One can see that many investigations have used the pairwise temperaturedifference estimator (Hand et al. 2012), which was inspired by the pairwise momentum estimator (Ferreira et al. 1999). However, in a different approach, recently Hill et al. (2016) and Ferraro et al. (2016) crosscorrelated the squared kSZ field with the WISE galaxy projected density catalogue and obtained a roughly 4 σ detection. This approach has similarities with the temperature dispersion that we probe in this paper, but is different in several ways. Firstly, in terms of tracers, Hill et al. (2016) used the density field of the WISEselected galaxies, of which only the radial distribution (W_{g}(η) kernel) is known, whereas in this work we use the MCXC Xray catalogue in which each cluster’s exact position and redshift are already determined. Secondly, for the optical depth treatment Hill et al. (2016) used 46 million WISE galaxies, so the crosscorrelation with the kSZ^{2} field contains the contribution of diffuse gas as well as virialized gas in groups; hence they assumed that their galaxies are tracing the velocity field with a uniform optical depth approximation (in their notation it is the f_{gas} parameter). In this paper, on the other hand, we are explicitly probing the velocity dispersion around galaxy clusters, so our optical depth comes from each individual cluster. Thirdly, in Hill et al. (2016) and Ferraro et al. (2016) the angular scales of the kSZ^{2} and WISE projected density field correlation lie in the range ℓ = 400–3000, whereas in this work, the dispersion effectively comes from a narrower range of scales around 5′–10′. Lastly, the lensing contamination in our case is below 1 σ of our signal, while in Hill et al. (2016) and Ferraro et al. (2016), the lensing is correlated with the WISE projected density field, and so the detected signal has a larger lensing contribution (Fig. 1 in Hill et al. 2016).
It is evident therefore that the methods discussed in this paper and described in Hill et al. (2016) are complementary. As the mapping of the smallscale CMB sky continues to improve, we can imagine that the assumptions made in either approach will need to be revisited and more sophisticated methods will be needed to probe the kSZ statistics more thoroughly, e.g., through direct comparison with simulations.
4. Implications for the peculiar velocity field
We now want to investigate what the temperature dispersion indicates for the variance of the peculiar velocity field. As shown in Eq. (1), the dimensionless temperature fluctuation is different from the dimensionless velocity field through the lineofsight optical depth factor τ. Since the coherence length of the velocity field is order 100 h^{−1} Mpc (Planck Collaboration Int. XXXVII 2016), i.e., much larger than the size of a cluster, the velocity can be taken out of the integral, giving(21)
In order to convert the kSZ signal into a lineofsight velocity we therefore need to obtain an estimate of τ for each cluster. In Planck Collaboration Int. XIII (2014), the values calculated are explicitly given as the optical depth per solid angle, obtained based on two scaling relations from Arnaud et al. (2005 and 2010). Here we adopt a slightly different approach, which is to determine the τ value at the central pixel of each galaxy cluster.
Many of the previous studies of the tSZ effect have used the “universal pressure profile” (UPP, Arnaud et al. 2010; Planck Collaboration Int. V 2013) and β isothermal model (Cavaliere & FuscoFemiano 1976, 1978) to model the pressure and electron density profiles of the clusters (Grego et al. 2000; Benson et al. 2003, 2004; Hallman et al. 2007; Halverson et al. 2009; Plagge et al. 2010). Because the UPP is just a fitting function for pressure, it is difficult to separate out the electron density and the temperature unless we use the isothermal assumption. In fact, Battaglia et al. (2012) demonstrated that the UPP is not absolutely universal, and that feedback from an active galactic nucleus can change the profile in a significant way. The functional form of the β model can be derived from a parameterization of density under the assumption of isothermality of the profile (e.g., Sarazin 1986). However, since isothermality is a poor assumption for many clusters (Planck Collaboration Int. V 2013), we only consider the β model here as a fitting function. Measurements of cluster profiles from the South Pole Telescope (SPT) have found that the index β = 0.86 provides the best fit to the profiles of SZ clusters (Plagge et al. 2010), and therefore we use this value of β in the following discussion.
The electron density can be written as(22)
where r_{c} = r_{vir}/c is the core radius of each cluster, with c being the concentration parameter. Here we adopt the formula from Duffy et al. (2008) and Komatsu et al. (2011) to calculate the concentration parameter given the redshift and halo mass of the cluster:(23)
In the catalogue, M_{500} and redshift z are given, so one can use these two quantities to calculate the virial mass M_{vir} of the cluster. The calculation is contained in Appendix B.
The radius r_{vir} is calculated through(24)
where ρ_{c}(z) is the critical density of the Universe at redshift z, and Δ(z) depends on Ω_{m} and Ω_{Λ} as Bryan & Norman (1998)(25)
with Ω(z) = Ω_{m}(1 + z)^{2}/[Ω_{m}(1 + z)^{3} Ω_{Λ}]. Thus,(26)
where Γ is the usual gamma function. To determine n_{e0}, we use , where(27)
Here the quantity f_{H} = 0.76 is the hydrogen mass fraction, m_{p} is the proton mass, and f_{gas} = (Ω_{b}/Ω_{m}) is the cosmic baryon fraction, while M_{500} is the cluster mass enclosed in the radius r_{500}. Thus,(28)
where c_{500} = r_{500}/r_{c} ≃ c_{vir}/2.0 is the concentration parameter for R_{500}.
Combining Eqs. (26), (27), and (28), we have(29)
In Fig. 13, we plot the histogram of the optical depth values of the 1526 clusters in the sample. The mean and standard deviation are given by τ = (3.9 ± 1.2) × 10^{−3}. This is very consistent with the quoted value τ = (3.75 ± 0.89) × 10^{−3} from the crosscorrelation between the kSZ SPT data and the photometric data from DES survey by fitting to a template of pairwise kSZ field (Soergel et al. 2016). Note that the uncertainty quoted here describes the scatter in the mean τ values for the whole of the sample.
We convert the temperature dispersion data listed in Table 5 to the lineofsight velocity dispersion measurement by using the modelled value of τ, after correcting by our estimate of the lensing effect. Our procedure is as follows. For each estimate of , we correct by the estimated shift caused by lensing, then calculate its υ^{2} value, and obtain an averaged value of υ^{2} via(30)
We then do this for the 5000 values of , and plot the distribution (after shifting by the lensing effect) in the right panel of Fig. 10 for the 2DILC map. We also present results for υ^{2} in Table 8, where we can see that for the conservative case, i.e., the 2DILC map, the velocity dispersion is measured to be υ^{2} = (12.3 ± 7.1) × (100 km s^{−1})^{2}. From the right panel of Fig. 10, we can see that the distribution is not completely Gaussian, but has a tail towards smaller υ^{2}. The frequency P(υ^{2} < 0) is 5.3%, which would correspond to a detection of the dispersion of peculiar velocity from 1526 MCXC clusters at the 1.7 σ level (using Appendix C).
Statistics of the lineofsight velocity dispersion υ^{2} (υ · r̂)^{2}.
In studies of peculiar velocity fields, the most relevant quantity is the linear lineofsight velocity (υ), or in other words 〈υ^{2}〉^{1/2}. We find^{7} 〈υ^{2}〉^{1/2} = (350 ± 100) km s^{−1} for the 2DILC map. One can see that the value we find is consistent with the velocity dispersion estimated through studies of the peculiar velocity field (e.g., Riess 2000; Turnbull et al. 2012; Ma & Scott 2013; Carrick et al. 2015).
Here we need to remember that what we measured is the lineofsight velocity dispersion, which contains both the largescale bulk flow, and the smallscale velocity and intrinsic dispersion (see, e.g., Ma & Scott 2014). The prediction for the rms bulk flow, Eq. (22) of Planck Collaboration Int. XIII (2014, or Eq. (4) in Ma & Pan 2014), is based on linear perturbation theory for the ΛCDM model and works only for the largescale bulk flows. The smallscale motions and intrinsic dispersion are not fully predictable from linear perturbation theory because they depend on subJeans scale structure evolution, which involves nonlinear effects. However, this smallscale velocity and intrinsic dispersion are nevertheless physical effects, which are nonnegligible in general (Carrick et al. 2015). One should consider that the lineofsight velocity dispersion that we have measured is a combination of two effects, namely largescale bulk flows and smallscale intrinsic dispersion, where the second component is generally nonnegligible.
We estimate that the histogram of separation distances between all pairs of cluster is peaked at d ≃ 600 Mpc. Since the bulk flow contributes to the velocity dispersion measurement here, then we can set an upper limit on the cosmic bulk flow on scales of 600 h^{−1} Mpc, (95% CL).
Such a constraint on largescale bulk flows indicates that the Universe is statistically homogeneous on scales of 600 h^{−1} Mpc. This is consistent with the limits obtained from TypeIa supernovae (Feindt et al. 2013), the Spiral Field Iband survey (Nusser & Davis 2011; Ma & Scott 2013), ROSAT galaxy clusters (Mody & Hajian 2012), and the Planck peculiar velocity study (Planck Collaboration Int. XIII 2014). However, it does not allow the very large “dark flow” claimed in Kashlinsky et al. (2008, 2010, 2012) and AtrioBarandela et al. (2015). In addition to ruling out such models, improved measurements of the velocity dispersion in the future have the potential to set up interesting constraints on dark energy and modified gravity (Bhattacharya & Kosowsky 2007, 2008).
5. Conclusions
The kinetic SunyaevZeldovich effect gives anisotropic perturbations of the CMB sky, particularly in the direction of clusters of galaxies. Previous studies have detected the kSZ effect through the pairwise momentum estimator and temperaturevelocity crosscorrelation. In this paper, we have detected the kSZ effect through a measurement of the temperature dispersion and then we have interpreted this as a determination of the smallscale velocity dispersion of cosmological structure.
To do this, we first selected two sets of Planck foregroundcleaned maps. One set contains four Planck publicly available maps, namely SMICA, NILC, SEVEM and Commander, each being produced using a different algorithm to minimize foreground emission. The second set, is the Planck2DILC map, which nulls the tSZ component, while resulting in slightly larger residual noise in the map. We then apply a matchedfilter technique to the maps, to suppress the primary CMB and instrumental noise. We specifically consider the MCXC cluster sample. Applying a Galactic and pointsource mask to the maps, results in 1526 MCXC clusters remaining unmasked.
We measured the distribution of the ΔT/T values for the 1526 MCXC cluster positions, and also at 1526 randomly selected positions, to give a quantification of the noise level. We found that the 1526 true cluster positions give extra variance to the distribution, and identify this as being due to the kSZ temperature dispersion effect. We compare this signal to results from 5000 random catalogues on the sky, each composed of 1526 random positions. This extra dispersion signal is persistent in several tests that we carry out.
We then construct estimators to quantify this effect. For the SMICA, NILC, and SEVEM maps, the significance of detection is stronger than in the 2DILC map, which is likely due to the fact that the residual tSZ effect in the map is correlated with the kSZ signal. In addition, an analytical estimate, supported by simulations, shows that about 0.7 σ of the temperaturedispersion effect comes from gravitational lensing. By subtracting this effect, quoting the conservative result from 2DILC, we obtain 〈s^{2}〉 = (1.35 ± 0.48) × 10^{−11} (68% CL), where . This gives a detection of temperature dispersion at about the 2.8 σ level. We obtain largely consistent results when we obtain results by weighting clusters with their different observed properties.
We further estimate the optical depth of each cluster, and thereby convert our (lensingcorrected) temperature dispersion measurement into a velocity dispersion measurement, obtaining 〈υ^{2}〉 = (12.3 ± 7.1) × (100 km s^{−1})^{2} (68% CL) using a Gaussian approximation. The distribution has P(υ^{2} < 0) = 5.3%, and the bestfit value is consistent with findings from largescale structure studies. This constraint implies that the Universe is statistically homogeneous on scales of 600 h^{−1} Mpc, with the bulk flow constrained to be (95% CL).
The measurement that we present here shows the promise of statistical kSZ studies for constraining the growth of structure in the Universe. To improve the results in the future, one needs to have better componentseparation algorithms to downweight the residual noise contained in the kSZ map, as well as having more sensitive and higher resolution CMB maps for removing the tSZ signal. One also needs larger cluster catalogues, with the uncertainty scaling roughly as if the residual noise is Gaussian.
Although in principle one could still have an isocurvature perturbation on very large scales (Turner 1991; Ma et al. 2011).
Planck (http://www.esa.int/Planck) is a project of the European Space Agency (ESA) with instruments provided by two scientific consortia funded by ESA member states and led by Principal Investigators from France and Italy, telescope reflectors provided through a collaboration between ESA and a scientific consortium led and funded by Denmark, and additional contributions from NASA (USA).
From the Planck Legacy Archive, http://pla.esac.esa.int
Available at https://github.com/carronj/LensIt
Acknowledgments
The Planck Collaboration acknowledges the support of: ESA; CNES, and CNRS/INSUIN2P3INP (France); ASI, CNR, and INAF (Italy); NASA and DoE (USA); STFC and UKSA (UK); CSIC, MINECO, JA, and RES (Spain); Tekes, AoF, and CSC (Finland); DLR and MPG (Germany); CSA (Canada); DTU Space (Denmark); SER/SSO (Switzerland); RCN (Norway); SFI (Ireland); FCT/MCTES (Portugal); ERC and PRACE (EU). A description of the Planck Collaboration and a list of its members, indicating which technical or scientific activities they have been involved in, can be found at http://www.cosmos.esa.int/web/planck/planckcollaboration. This paper makes use of the HEALPix software package.
References
 Aghanim, N., Górski, K. M., & Puget, J.L. 2001, A&A, 374, 1 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
 Arnaud, M., Pointecouteau, E., & Pratt, G. W. 2005, A&A, 441, 893 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
 Arnaud, M., Pratt, G. W., Piffaretti, R., et al. 2010, A&A, 517, A92 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
 AtrioBarandela, F., Kashlinsky, A., Ebeling, H., Fixsen, D. J., & Kocevski, D. 2015, ApJ, 810, 143 [NASA ADS] [CrossRef] [Google Scholar]
 Basak, S., & Delabrouille, J. 2012, MNRAS, 419, 1163 [NASA ADS] [CrossRef] [Google Scholar]
 Basak, S., & Delabrouille, J. 2013, MNRAS, 435, 18 [NASA ADS] [CrossRef] [Google Scholar]
 Battaglia, N., Bond, J. R., Pfrommer, C., & Sievers, J. L. 2012, ApJ, 758, 74 [NASA ADS] [CrossRef] [Google Scholar]
 Benson, B. A., Church, S. E., Ade, P. A. R., et al. 2003, ApJ, 592, 674 [NASA ADS] [CrossRef] [Google Scholar]
 Benson, B. A., Church, S. E., Ade, P. A. R., et al. 2004, ApJ, 617, 829 [NASA ADS] [CrossRef] [Google Scholar]
 Bhattacharya, S., & Kosowsky, A. 2007, ApJ, 659, L83 [NASA ADS] [CrossRef] [Google Scholar]
 Bhattacharya, S., & Kosowsky, A. 2008, Phys. Rev. D, 77, 083004 [NASA ADS] [CrossRef] [Google Scholar]
 Böhringer, H., Voges, W., Huchra, J. P., et al. 2000, ApJ, 129, 435 [Google Scholar]
 Böhringer, H., Schuecker, P., Guzzo, L., et al. 2004, A&A, 425, 367 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
 Bryan, G. L., & Norman, M. L. 1998, ApJ, 495, 80 [NASA ADS] [CrossRef] [Google Scholar]
 Burenin, R. A., Vikhlinin, A., Hornstrup, A., et al. 2007, ApJ, 172, 561 [NASA ADS] [Google Scholar]
 Burke, D. J., Collins, C. A., Sharples, R. M., Romer, A. K., & Nichol, R. C. 2003, MNRAS, 341, 1093 [NASA ADS] [CrossRef] [Google Scholar]
 Carrick, J., Turnbull, S. J., Lavaux, G., & Hudson, M. J. 2015, MNRAS, 450, 317 [NASA ADS] [CrossRef] [Google Scholar]
 Cavaliere, A., & FuscoFemiano, R. 1976, A&A, 49, 137 [NASA ADS] [Google Scholar]
 Cavaliere, A., & FuscoFemiano, R. 1978, A&A, 70, 677 [NASA ADS] [Google Scholar]
 Cruddace, R., Voges, W., Böhringer, H., et al. 2002, ApJ, 140, 239 [NASA ADS] [Google Scholar]
 Davies, R. L., Efstathiou, G., Fall, S. M., Illingworth, G., & Schechter, P. L. 1983, ApJ, 266, 41 [NASA ADS] [CrossRef] [Google Scholar]
 De Bernardis, F., Aiola, S., Vavagiakis, E. M., et al. 2017, J. Cosmol. Astropart. Phys., 3, 008 [Google Scholar]
 Delabrouille, J., Cardoso, J.F., Le Jeune, M., et al. 2009, A&A, 493, 835 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
 Dodelson, S. 2004, Phys. Rev. D, 70, 023009 [NASA ADS] [CrossRef] [Google Scholar]
 Duffy, A. R., Schaye, J., Kay, S. T., & Dalla Vecchia, C. 2008, MNRAS, 390, L64 [NASA ADS] [CrossRef] [Google Scholar]
 Ebeling, H., Edge, A. C., Bohringer, H., et al. 1998, MNRAS, 301, 881 [NASA ADS] [CrossRef] [Google Scholar]
 Ebeling, H., Edge, A. C., Allen, S. W., et al. 2000, MNRAS, 318, 333 [NASA ADS] [CrossRef] [Google Scholar]
 Ebeling, H., Barrett, E., Donovan, D., et al. 2007, ApJ, 661, L33 [NASA ADS] [CrossRef] [Google Scholar]
 Ebeling, H., Edge, A. C., Mantz, A., et al. 2010, MNRAS, 407, 83 [NASA ADS] [CrossRef] [Google Scholar]
 Feindt, U., Kerschhaggl, M., Kowalski, M., et al. 2013, A&A, 560, A90 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
 Feldman, H. A., Watkins, R., & Hudson, M. J. 2010, MNRAS, 407, 2328 [NASA ADS] [CrossRef] [Google Scholar]
 Ferraro, S., Hill, J. C., Battaglia, N., Liu, J., & Spergel, D. N. 2016, Phys. Rev. D, 94, 123526 [NASA ADS] [CrossRef] [Google Scholar]
 Ferreira, P. G., Juszkiewicz, R., Feldman, H. A., Davis, M., & Jaffe, A. H. 1999, ApJ, 515, L1 [NASA ADS] [CrossRef] [Google Scholar]
 Gioia, I. M., & Luppino, G. A. 1994, ApJ, 94, 583 [NASA ADS] [CrossRef] [Google Scholar]
 Górski, K. M., Hivon, E., Banday, A. J., et al. 2005, ApJ, 622, 759 [NASA ADS] [CrossRef] [Google Scholar]
 Grego, L., Carlstrom, J. E., Joy, M. K., et al. 2000, ApJ, 539, 39 [NASA ADS] [CrossRef] [Google Scholar]
 Guilloux, F., Faÿ, G., & Cardoso, J.F. 2009, Appl. Comput. Harmon. Anal., 26, 143 [CrossRef] [Google Scholar]
 Haehnelt, M. G., & Tegmark, M. 1996, MNRAS, 279, 545 [NASA ADS] [CrossRef] [Google Scholar]
 Hallman, E. J., Burns, J. O., Motl, P. M., & Norman, M. L. 2007, ApJ, 665, 911 [NASA ADS] [CrossRef] [Google Scholar]
 Halverson, N. W., Lanting, T., Ade, P. A. R., et al. 2009, ApJ, 701, 42 [NASA ADS] [CrossRef] [Google Scholar]
 Hand, N., Addison, G. E., Aubourg, E., et al. 2012, Phys. Rev. Lett., 109, 041101 [NASA ADS] [CrossRef] [Google Scholar]
 Henry, J. P. 2004, ApJ, 609, 603 [NASA ADS] [CrossRef] [Google Scholar]
 Henry, J. P., Mullis, C. R., Voges, W., et al. 2006, ApJ, 162, 304 [NASA ADS] [Google Scholar]
 HernándezMonteagudo, C., Ma, Y.Z., Kitaura, F. S., et al. 2015, Phys. Rev. Lett., 115, 191301, [NASA ADS] [CrossRef] [Google Scholar]
 Hill, J. C., Ferraro, S., Battaglia, N., Liu, J., & Spergel, D. N. 2016, Phys. Rev. Lett., 117, 051301 [NASA ADS] [CrossRef] [Google Scholar]
 Hivon, E., Górski, K. M., Netterfield, C. B., et al. 2002, ApJ, 567, 2 [NASA ADS] [CrossRef] [Google Scholar]
 Holzapfel, W. L., Ade, P. A. R., Church, S. E., et al. 1997, ApJ, 481, 35 [NASA ADS] [CrossRef] [Google Scholar]
 Horner, D. J., Perlman, E. S., Ebeling, H., et al. 2008, ApJ, 176, 374 [NASA ADS] [Google Scholar]
 Kashlinsky, A., & AtrioBarandela, F. 2000, ApJ, 536, L67 [NASA ADS] [CrossRef] [Google Scholar]
 Kashlinsky, A., AtrioBarandela, F., Kocevski, D., & Ebeling, H. 2008, ApJ, 686, L49 [NASA ADS] [CrossRef] [Google Scholar]
 Kashlinsky, A., AtrioBarandela, F., Kocevski, D., & Ebeling, H. 2009, ApJ, 691, 1479 [NASA ADS] [CrossRef] [Google Scholar]
 Kashlinsky, A., AtrioBarandela, F., Ebeling, H., Edge, A., & Kocevski, D. 2010, ApJ, 712, L81 [NASA ADS] [CrossRef] [Google Scholar]
 Kashlinsky, A., AtrioBarandela, F., & Ebeling, H. 2012, ArXiv eprints [arXiv: 1202.0717] [Google Scholar]
 Kocevski, D. D., Ebeling, H., Mullis, C. R., & Tully, R. B. 2007, ApJ, 662, 224 [NASA ADS] [CrossRef] [Google Scholar]
 Komatsu, E., Smith, K. M., Dunkley, J., et al. 2011, ApJ, 192, 18 [NASA ADS] [Google Scholar]
 Kormendy, J. 2001, ASP Conf. Ser., 230, 247 [NASA ADS] [Google Scholar]
 Kormendy, J., & Bender, R. 1996, ApJ, 464, L119 [NASA ADS] [CrossRef] [Google Scholar]
 Lewis, A., & Challinor, A. 2006, Phys. Rept., 429, 1 [Google Scholar]
 Li, Y.C., Ma, Y.Z., Remazeilles, M., & Moodley, K. 2018, Phys. Rev. D, 97, 023514 [NASA ADS] [CrossRef] [Google Scholar]
 Ma, Y.Z., & Pan, J. 2014, MNRAS, 437, 1996 [NASA ADS] [CrossRef] [Google Scholar]
 Ma, Y.Z., & Scott, D. 2013, MNRAS, 428, 2017 [NASA ADS] [CrossRef] [Google Scholar]
 Ma, Y.Z., & Scott, D. 2014, Astron. Geophys., 55, 030003 [CrossRef] [Google Scholar]
 Ma, Y.Z., Gordon, C., & Feldman, H. A. 2011, Phys. Rev. D, 83, 103002 [NASA ADS] [CrossRef] [Google Scholar]
 Ma, Y.Z., Hinshaw, G., & Scott, D. 2013, ApJ, 771, 137 [NASA ADS] [CrossRef] [Google Scholar]
 MacMillan, J. D., Widrow, L. M., & Henriksen, R. N. 2006, ApJ, 653, 43 [NASA ADS] [CrossRef] [Google Scholar]
 Mody, K., & Hajian, A. 2012, ApJ, 758, 4 [NASA ADS] [CrossRef] [Google Scholar]
 Mullis, C. R., McNamara, B. R., Quintana, H., et al. 2003, ApJ, 594, 154 [NASA ADS] [CrossRef] [Google Scholar]
 Narcowich, F., Petrushev, P., & Ward, J. 2006, SIAM J. Math. Anal., 38, 574 [CrossRef] [Google Scholar]
 Navarro, J. F., Frenk, C. S., & White, S. D. M. 1996, ApJ, 462, 563 [NASA ADS] [CrossRef] [Google Scholar]
 Nusser, A., & Davis, M. 2011, ApJ, 736, 93 [NASA ADS] [CrossRef] [Google Scholar]
 Ostriker, J. P. 1980, Comments Astrophys., 8, 177 [NASA ADS] [Google Scholar]
 Peebles, P. J. E. 1980, in The largescale structure of the universe (Research supported by the National Science Foundation) (Princeton, NJ: Princeton University Press) [Google Scholar]
 Perlman, E. S., Horner, D. J., Jones, L. R., et al. 2002, ApJ, 140, 265 [Google Scholar]
 Piffaretti, R., Arnaud, M., Pratt, G. W., Pointecouteau, E., & Melin, J.B. 2011, A&A, 534, A109 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
 Plagge, T., Benson, B. A., Ade, P. A. R., et al. 2010, ApJ, 716, 1118 [NASA ADS] [CrossRef] [Google Scholar]
 Planck Collaboration XII. 2014, A&A, 571, A12 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
 Planck Collaboration IX. 2016, A&A, 594, A9 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
 Planck Collaboration X. 2016, A&A, 594, A10 [Google Scholar]
 Planck Collaboration XIII. 2016, A&A, 594, A13 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
 Planck Collaboration XXVI. 2016, A&A, 594, A26 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
 Planck Collaboration Int. V. 2013, A&A, 550, A13 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
 Planck Collaboration Int. XIII. 2014, A&A, 561, A97 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
 Planck Collaboration Int. XXXVII. 2016, A&A, 586, A14 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
 Planck Collaboration Int. XLIII. 2016, A&A, 596, A104 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
 Remazeilles, M., Delabrouille, J., & Cardoso, J.F. 2011a, MNRAS, 410, 2481 [NASA ADS] [CrossRef] [Google Scholar]
 Remazeilles, M., Delabrouille, J., & Cardoso, J.F. 2011b, MNRAS, 418, 467 [NASA ADS] [CrossRef] [Google Scholar]
 Remazeilles, M., Dickinson, C., Banday, A. J., BigotSazy, M.A., & Ghosh, T. 2015, MNRAS, 451, 4311 [NASA ADS] [CrossRef] [Google Scholar]
 Riess, A. G. 2000, ASP Conf. Ser., 201, 80 [NASA ADS] [Google Scholar]
 Romer, A. K., Nichol, R. C., Holden, B. P., et al. 2000, ApJ, 126, 209 [NASA ADS] [Google Scholar]
 Sarazin, C. L. 1986, Rev. Mod. Phys., 58, 1 [NASA ADS] [CrossRef] [Google Scholar]
 Schaan, E., Ferraro, S., VargasMagaña, M., et al. 2016, Phys. Rev. D, 93, 082002 [NASA ADS] [CrossRef] [Google Scholar]
 Soergel, B., Flender, S., Story, K. T., et al. 2016, MNRAS, 461, 3172 [NASA ADS] [CrossRef] [Google Scholar]
 Sugiyama, N. S., Okumura, T., & Spergel, D. N. 2018, MNRAS, 475, 3764 [NASA ADS] [CrossRef] [Google Scholar]
 Sunyaev, R. A., & Zeldovich, Y. B. 1972, Comments Astrophys. Space Phys., 4, 173 [NASA ADS] [EDP Sciences] [Google Scholar]
 Sunyaev, R. A., & Zeldovich, I. B. 1980, MNRAS, 190, 413 [NASA ADS] [CrossRef] [Google Scholar]
 Tegmark, M., & de OliveiraCosta, A. 1998, ApJ, 500, L83 [NASA ADS] [CrossRef] [Google Scholar]
 Turnbull, S. J., Hudson, M. J., Feldman, H. A., et al. 2012, MNRAS, 420, 447 [NASA ADS] [CrossRef] [Google Scholar]
 Turner, M. S. 1991, Phys. Rev. D, 44, 3737 [NASA ADS] [CrossRef] [Google Scholar]
 Waelkens, A., Maturi, M., & Enßlin, T. 2008, MNRAS, 383, 1425 [NASA ADS] [CrossRef] [Google Scholar]
 Watkins, R., Feldman, H. A., & Hudson, M. J. 2009, MNRAS, 392, 743 [NASA ADS] [CrossRef] [Google Scholar]
Appendix A: The statistics of s^{2}
Let us first define the kth moment of a distribution of a random variable x to be(A.1)
The estimator is defined in Eq. (16). Note that the observed is always taken to be the value of kSZ on the true cluster position, so there is no randomness in . We also define(A.2)
Therefore, the mean value of is(A.3)
Therefore, we first calculate(A.5)
where in the above derivation we have assumed that the residual noise samples in two different pixels are uncorrelated, i.e., 〈n_{i}n_{j}〉 = 0. Finally,(A.7)
Appendix B: Converting M_{500} to M_{vir}
For each cluster, M_{500} is defined as the mass within the radius of R_{500}, in which its average density is 500 times the critical density of the Universe,(B.1)
and E^{2}(z) = Ω(1+z)^{3} + Ω_{Λ}. The quantity M_{vir} is calculated via Eqs. (24) and (25), and the relationship between M_{500} and M_{vir }is Mody & Hajian (2012)(B.3)
where c is the concentration parameter (Eq. (23)) and m(x) = ln(1 + x) − x/(1 + x). Given redshift z and mass M_{500}, we can thus determine r_{500} through Eq. (B.1). If we substitute M_{500}, z, and r_{500} into Eq. (B.3), this becomes an algebraic equation for M_{vir}. This is because r_{vir} can be determined from M_{vir} through Eqs. (24) and (25), and c is related to M_{vir} through Eq. (23). Therefore, we can iteratively solve for M_{vir}, given the values of M_{500} and z.
Appendix C: Converting P values into S/Ns
Since the distribution of weighted s^{2} has a longer tail than a Gaussian distribution, instead of calculating the ratio between mean and rms values of the distribution, we calculate the pvalues, and list them in the third column of Table 6. We now convert them into S/N in the following way. Suppose the variable x satisfies a Gaussian distribution, the normalized distribution is . Then the cumulative probability to find x < 0 is(C.1)
is the complimentary error function.
Therefore the equivalent S/N given the value of is(C.3)
We use Eq. (C.3) to obtain the fourth column of Table 6.
All Tables
Recent measurements of the kinetic SunyaevZeldovich effect with crosscorrelations of various tracers of largescale structure.
Statistics of the values of (ΔT/T) × 10^{5} at the true 1526 cluster positions and for 1526 randomlyselected positions.
rms values for the true sky positions of 1526 MCXC catalogue clusters (σ_{MCXC}), along with the mean (σ_{ran}) and scatter (σ(σ_{ran})) of the values of the rms for 5000 random catalogues, where each catalogue consists of 1526 random positions on the sky.
Statistics of the weighted variables for different choices of weights in the 2DILC map.
Statistics of split samples. Here the median values of the samples are (M_{500})_{mid} = 1.68 × 10^{14}M_{⊙}, (L_{500})_{mid} = 2.3 × 10^{10}L_{⊙}, and (R_{500})_{mid} = 0.79 Mpc.
All Figures
Fig. 1.
Top left: stack of the NILC CMB map in the directions of Planck SZ (PSZ) galaxy clusters. Top right: stack of the 2DILC CMB map in the direction of PSZ galaxy clusters. This sample provides a very stringent test of the tSZ leakage, since the PSZ positions are the known places on the sky with detectable SZ signal. The stacked NILC CMB map clearly shows an excess in the centre, which is due to residual contamination from the tSZ effect, while the 2DILC CMB map has a signature in the centre that is consistent with the strength of other features in the stacked image. Bottom left: stack of the NILC CMB map in the directions of MCXC clusters. Bottom right: stack of the 2DILC CMB map in the direction of 1526 MCXC clusters (see Sect. 2.2 for the detail of the catalogue). For a different set of sky positions, the results are broadly consistent with those for the PSZ clusters. All these maps are 3° × 3° in size, and use the same colour scale. 

In the text 
Fig. 2.
Profiles for stacked patches (see Fig. 1) of the NILC CMB map (black diamonds) and the 2DILC CMB map (blue triangles) at the positions of PSZ clusters (top panel) and MCXC clusters (bottom panel). The profile of the stacked Planck 217GHz map is also shown as a reference (green squares). The central deficit in the flux profile of the stacked NILC CMB map (black diamonds) is due to residual tSZ contamination. 

In the text 
Fig. 3.
Left: fullsky distribution of 1526 MCXC Xray clusters (Piffaretti et al. 2011; Planck Collaboration Int. XIII 2014) in Galactic coordinates. The dark blue area is the masked region, and the clusters are shown in orange. Right: redshift histogram of 1526 Xray clusters, with bin width Δz = 0.025. 

In the text 
Fig. 4.
Measured (black dots) and predicted (red line) power spectra from the Planck2DILC map. The predicted spectrum is based on the bestfitting ΛCDM model convolved with the squared beam , with the noise added. These are estimated using the pseudoC_{ℓ} estimator described in Hivon et al. (2002). 

In the text 
Fig. 5.
Optimal matched filter (black line) for pointsource detection in the Planck2DILC map (Eq. (14)). For comparison, the power spectra of the CMB signal (red line) and noise map (blue dashed line) are shown, along with their sum (brown line). 

In the text 
Fig. 6.
Filtered (with Eq. (14)) and masked 2DILC map in dimensionless units (i.e., ΔT/T). 

In the text 
Fig. 7.
Histograms of 1526 ΔT/T values of 2DILC map at the cluster catalogue positions (red bars), and randomly selected positions (black bars). The statistics of the true cluster positions and random positions can be found in Table 2. 

In the text 
Fig. 8.
Histogram of the values of 5000 random catalogues on the sky (each having N_{c} = 1526), with different colours representing different Planck foregroundcleaned maps. The 68% width of the 2DILC, SMICA, NILC, SEVEM and Commander histograms are 2.86, 2.49, 2.48, 2.53, 2.62 (×10^{−7}), respectively. The vertical lines represent the average ΔT/T values at true MCXC cluster positions for each map. One can see that only for the 2DILC map is this value within the 68% range of the random catalogue distribution, while others are quite far off. This indicates that, except for the 2DILC map, the public Planck maps have residual tSZ contamination at the cluster positions. 

In the text 
Fig. 9.
Distribution of the rms for 5000 mock catalogues (yellow histogram); here each catalogue consists of N_{c} randomly chosen positions on the filtered 2DILC map. The mean and rms of the 5000 random catalogues are 1.10 × 10^{−5} (shown as the black dashed vertical line) and 2.15 × 10^{−7}, respectively. For the N_{c} true MCXC cluster positions, the rms is 1.17 × 10^{−5} (red vertical dashed line). 

In the text 
Fig. 10.
Left: distribution of 5000 values of s^{2} with uniform weight (Eq. (16)) on each position for the 2DILC map. Right: distribution of υ^{2} calculated from Eq. (30), after subtracting the lensing shift. The distribution has P(υ^{2} < 0) = 5.3%, corresponding to a (1sided) detection of 1.6 σ. We have tested with 50 000 values of s^{2} and the results are consistent with those from 5000 values. 

In the text 
Fig. 11.
Left: R_{500} versus M_{500} for the sample. Right: L_{500} versus M_{500} for the sample. The vertical and horizontal lines show the median values of R_{500}, M_{500}, and L_{500}. 

In the text 
Fig. 12.
Histogram of the values of 5000 random catalogues (each having N_{c} = 1526) on simulated lensed skies with added 2DILC noise. The mean and rms for the 5000 mock catalogues are 1.07 × 10^{−5} and 2.13 × 10^{−7}, respectively. The vertical dashed lines indicate the values at the MCXC cluster positions before and after the clusterlensing effect is added; the values are 1.047 × 10^{−5} before (purple line) and 1.061 × 10^{−5} after (red line). 

In the text 
Fig. 13.
Histogram of the optical depth τ derived using Eq. (29) for 1526 Xray cluster positions. 

In the text 
Current usage metrics show cumulative count of Article Views (fulltext article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 4896 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.