Large Scale Distribution of Galaxies

Home , Norma (constellation), Norma Cluster

Large scale distribution of galaxies

Projections on the sky Pie diagrams Motion of galaxies Constrained simulations Voids Galaxies: effects of environment Projection

1986ApJ...302L...1D

1986ApJ...302L...1D 1986ApJ...302L...1D 1986ApJ...302L...1D 1986ApJ...302L...1D 168 O. Lahav et al.

For comparison, we also apply our analysis to the complete which the density of galaxies is enhanced can be found by IRAS 1.2-Jy sample (Fisher et al. 1995a; Strauss et al. 1992), calculating the covariance matrix: I x^ x^ , where the 168 O. Lahav et al. ij gal i j which includes 5313 galaxies over 87.6 per cent of the sky. The Cartesian axes xÃi i 1; 2; 3 are defined for a unit sphere. The P ZOA in IRAS was interpolated according to the YahilForet al.comparison,(1991) matrixwe alsoIij canapplybe diagonalizedour analysis(cf. Sectionto 4.2)the andcompletethe smallest which the density of galaxies is enhanced can be found by procedure mentioned above. Hereafter, we refer to this sample as eigenvector indicates the direction of the normal to the great the IRAS sample. We note that the ORS and IRASIRASsamples1.2-Jyare notsamplecircle.(FisherWe consideredet al.the1995a;projectedStraussIRAS galaxyet distributional. 1992),out to calculating the covariance matrix: Iij galx^ix^j, where the 21 really independent; roughly 60 per cent of whichthe IRASincludes1.2-Jy 531340 andgalaxies80 h Mpc,overand87.6foundperthatcentthe greatof circlethe skis alignedy. Thewith Cartesian axes xÃi i 1; 2; 3 are defined for a unit sphere. The 21 galaxies out to 80 h Mpc are also cataloguedZOAinintheIRASORSwas interpolatedthe standard deaccordingVaucouleurs'to theSGPYahilto withinet al.19(1991)8 and 78, matrix I can be diagonalized (cf. Section P4.2) and the smallest magnitude-limited sample. respectively. Can we see the translation of the great circle to a ij We work purely in Local Group redshiftprocedurespace, exceptmentionedin planeabovin threee. Hereafterdimensions?, weFig.refer1 shotowsthisthesampledistributionas of eigenvector indicates the direction of the normal to the great Section 6 on Wiener reconstruction. the IRAS sample. Wgalaxiese noteprojectedthat theonORSthe skyandin theIRASIRASsamplesand ORS samples.are notThe circle. We considered the projected IRAS galaxy distribution out to really independent;distributionroughlyis highly60 pernon-isotropic,cent of andthetheIRASSGP is 1.2-Jyclearly the 40 and 80 h21 Mpc, and found that the great circle is aligned with most2prominent1 feature. We note that the structures seen in ORS 3 A B R I E F T O U R O F T H E S G P galaxies out to 80andh IRASMpcare quiteare similaralso , cataloguedbut the ORS mapin isthemoreORSdensely the standard de Vaucouleurs' SGP to within 198 and 78, magnitude-limited sampled,sample.and clusters are more prominent. Fig. 2 shows the raw respectively. Can we see the translation of the great circle to a In this section, we first show the distribution of galaxies in the Mon. Not. R. Astron. Soc. 312, 166±176 (2000) distributions, of ORS and IRAS galaxies, uncorrected for selection samples projected on to the various supergalactic planes.We workWe thenpurely in Local Group redshift space, except in plane in three dimensions? Fig. 1 shows the distribution of effects, in a sphere of radius of 40 h21 Mpc projected on the quantify this by calculating the density field in shells as a function Section 6 on Wienerstandardreconstruction.(SGX±SGZ ), (SGX±SGY ) and (SGY±SGZ ) planes. The galaxies projected on the sky in the IRAS and ORS samples. The of SGZ. Finally, we quantify the centre of mass of the overdensity. The supergalactic plane revisited with theSGPOpticalis particularlyRedshiftvisible edge-onSurveyin the (SGX±SGZ ) and distribution is highly non-isotropic, and the SGP is clearly the (SGY±SGZ ) projections. most prominent feature. We note that the structures seen in ORS Maps of the smoothed density field, after applying the 3.1 Visual impr1,2wession 3 1 4 5 and IRAS are quite similar, but the ORS map is more densely O. Lahav, B. X. Santiago, A.3 M.A WB RebsterI E F, TMichaelappropriateO U R OA.weights,F TStrauss,HareE shoS Gwn²PMelse. whereDavis,for the IRAS survey Historically, the6SGP was identified in projected7 2D maps. We (Saunders et al. 1991; Strauss et al. 1992; Fisher et al. 1995b; sampled, and clusters are more prominent. Fig. 2 shows the raw A.usedDresslerthe IRAS dataand(whichJ.isPuniformly. HuchrasampledInovthiser thesection,sky) to weStraussfirst &shoWwillickthe1995;distributionWebster, Lahavof &galaxiesFisher 1997)in theand for 1Institute of Astronomy, Madingley Road, Cambridge CB3 0HA distributions, of ORS and IRAS galaxies, uncorrected for selection revisit the 2D appearance of the SGP by fittingsamplesthe data to projecteda great onORSto(Santiagothe variouset al. 1995,supergal1996;acticBakerplanes.et al. 1998).We InthenFigs 3 21 2Racah Institute of Physics, The Hebrew University, Jerusalem 91904, Israel circle. For a projected distribution of sources the great circle along and 4 we show histograms of the galaxy density field d, corrected effects, in a sphere of radius of 40 h Mpc projected on the 3Instituto de FõÂsica, Universidade Federal do Rio Grandquantifye do Sul, 91501-970,this byPortocalculatingAlegre, RS, Braztheil density field in shells as a function 4 standard (SGX±SGZ ), (SGX±SGY ) and (SGY±SGZ ) planes. The Princeton University Observatory, Princeton, NJ 08544,ofUSASGZ. Finally, we quantify the centre of mass of the overdensity. 5Physics and Astronomy Departments, University of California, Berkeley, CA 94720, USA SGP is particularly visible edge-on in the (SGX±SGZ ) and 6Observatories of the Carnegie Institution, 813 Santa Barbara St, Pasadena, CA 91101, USA 7Center for Astrophysics, 60 Garden St, Cambridge, MA 02138, USA (SGY±SGZ ) projections. Maps of the smoothed density field, after applying the 3.1 Visual impression appropriate weights, are shown elsewhere for the IRAS survey Accepted 1999 September 20. Received 1999 September 6; in original form 1998 October 5 Historically, the SGP was identified in projected 2D maps. We (Saunders et al. 1991; Strauss et al. 1992; Fisher et al. 1995b; used the IRAS data (which is uniformly sampled over the sky) to Strauss & Willick 1995; Webster, Lahav & Fisher 1997) and for A B S T R A CreTvisit the 2D appearance of the SGP by fitting the data to a great ORS (Santiago et al. 1995, 1996; Baker et al. 1998). In Figs 3 We re-examine the existence and extent of the planar structure in the local galaxy density field, the so-calledcircle. supergalacticFor a projectedplanedistrib(SGP).utionThis ofstructuresourcesis studtheiedgreatherecirclein threealong and 4 we show histograms of the galaxy density field d, corrected dimensions using both the new Optical Redshift Survey (ORS) and the IRAS 1.2-Jy redshift survey. The density contrast in a slab of thickness 20 h21 Mpc and diameter 80 Mpc aligned with the standard de Vaucouleurs supergalactic coordinates is dsgp , 0:5 for both ORS and IRAS. The structure of the SGP is not well described by a homogeneous ellipsoid, although it does appear to be a flattened structure, which we quantify by calculating the moment of inertia tensor of the density field. The directions of the principal axes vary with radius, but the minor axis remains within uz , 308 of the standard SGP Z-axis, out to a radius of 80 h21 Mpc, for both ORS and IRAS. However, the structure changes shape with radius, varying between a flattened pancake and a dumbbell, the latter at a radius of ,50 h21 Mpc, Figure 1. The distribution of galaxies projected on thewhereskytheinGrtheeat AttractorIRAS andandORSPerseussamples.±Pisces superclThisustersis andominateAitofthef projectiondistribution. inThissupergalactic coordinates, with 21 SGL 908; SGB 0 (close to the Virgo cluster) in the callscentreto questioof thenmap.the connectiObjectsvity withinof the `plane'2000beyondkm s 40areh2sho1 Mpc.wnTheas circledconfigurationcrosses; objects between 2000 and , 4000 km s21 are indicated as crosses, and dots mark thefoundpositionshere canofbemorevieweddistantas part ofobjects.a web ofHerefilamentwes andincludesheets,onlyrathercataloguedthan as an isolatedgalaxies, which is why the zone of avoidance is so prominent in these two figures. pancake-like structure. An optimal minimum variance reconstruction of the density field using Wiener filtering, which corrects for both redshift distortion and shot noise, yields a similar misalignment angle and behaviour of axes. The background-independent statistic of axes proposed here can be best used for testing cosmological models by comparisonq 2000with RAS, MNRAS 312, 166±176 N-body simulations. Key words: methods: statistical ± galaxies: distances and redshifts ± large-scale structure of Figure 1. The distribution of galaxiesUniprojectedverse.on the sky in the IRAS and ORS samples. This is an Aitoff projection in supergalactic coordinates, with SGL 908; SGB 0 (close to the Virgo cluster) in the centre of the map. Objects within 2000 km s21 are shown as circled crosses; objects between 2000 and 4000 km s21 are indicated as crosses, and dots mark the positions of more distant objects. Here we include only catalogued galaxies, which is why the zone of structure in the galaxy distribution. This should be distinguished 1avoidanceI N T RisOsoDprominentU C T I O inN these two figures. from the formal definition of supergalactic coordinates. de The major planar structure in the local universe, the supergalactic Vaucouleurs, de Vaucouleurs & Corwinq(12000976) RAS,and deMNRASVaucoul312eurs, 166±176 plane (SGP), was recognized by de Vaucouleurs (1953, 1956, et al. (1991) defined a spherical-coordinate system SGL, SGB in 1958, 1975a,b) using the Shapley±Ames catalogue, following an which the equator roughly lies along the SGP (as identified at the earlier analysis of radial velocities of nearby galaxies, which time), with the North Pole SGB 908 in the direction of suggested a differential rotation of the `metagalaxy' by Rubin Galactic coordinates (l 47:378; b 6:328). The position (1951). This remarkable feature in the distribution of nebulae was SGL 08; SGB 08 is at (l 137:378; b 08), which is one of in fact already noticed by William Herschel more than 200 years the two regions where the SGP is crossed by the Galactic plane. earlier (for historical reviews see Flin 1986, Rubin 1989). The Virgo cluster lies in the direction (SGL 1048; SGB 228). When referring to the SGP in this paper we mean the planar Traditionally, the Virgo cluster was regarded as the centre of an overdense region called the `Local Supercluster' (e.g. Davis & w E-mail: [email protected] Huchra 1982; Hoffman & Salpeter 1982; Tully & Shaya 1984; ² Alfred P. Sloan Foundation Fellow, and Cottrell Fellow of Research Lilje, Yahil & Jones 1986). However, some of the much larger Corporation. superclusters that are seen in recent redshift surveys, such as the

q 2000 RAS

Figure 1. The distribution of galaxies projected on the sky in the IRAS and ORS samples. This is an Aitoff projection in supergalactic coordinates, with SGL 908; SGB 0 (close to the Virgo cluster) in the centre of the map. Objects within 2000 km s21 are shown as circled crosses; objects between 2000 and 4000 km s21 are indicated as crosses, and dots mark the positions of more distant objects. Here we include only catalogued galaxies, which is why the zone of avoidance is so prominent in these two figures.

q 2000 RAS, MNRAS 312, 166±176 The supergalactic plane 169

The supergalactic plane 169

Figure 3. Histograms of the ORS density field (corrected for selection effects) as a function of SGZ, averaged over SGX and SGY, within spheres of radius as indicated in each panel. The binning is 5 h21 Mpc. The error Figure 2. The distributions of ORS and IRAS galaxies in a sphere of radius bars are Poissonian. of 40 h21 Mpc, projected on the standard (SGX±SGZ ), (SGX±SGY ) and (SGY±SGZ ) planes. The SGP is visible edge-on as the linear feature at SGZ 0 in the (SGX±SGZ ) and (SGY±SGZ ) projections ORS includes IRAS galaxies (real and mock) in the zone of avoidance (see text). for selection effects, as a function of SGZ, averaged over SGX and SGY, within spheres of ever-larger radii, as indicated in each panel. The SGP is seen as an enhancement at SGZ 0 in these plots to roughly R 60 h21 Mpc, but is not apparent on larger scales. max

3.2 The level of overdensity Figure 4. Histograms of the IRAS density field (corrected for selection effects) as a function of SGZ, averaged over SGX and SGY, within spheres Our visual impression is that there is indeed a flattenedFigure 2.structureThe distributionsofofradiusORS andas indicatedIRAS galaxiesin eachin apanel.sphereTheof radiuserror bars are Poissonian. in the galaxy distribution, aligned along SGZ of0,40exteh21ndingMpc, projectedto on the standard (SGX±SGZ ), (SGX±SGY ) and 21 3 appreciable redshifts. To quantify this we calculate(SGYthe ±ovSGZerdensity) planes. The cubesSGP is ofvisiblevolumeedge-on(40ash theMpc)linear arefeaturedrmsat, 0:37 and ,0.30 for SGZ 0 in the (SGX±SGZoptical) and ((Stromlo/APM)SGY±SGZ ) projectiandonsIRASORS includes(1.2-Jy) surveys, respectively dsgp ; nsgp=knl 2 1 as a function of Rmax, where nsgp is evaluated within a circular cylinder centred on the Local GroupIRAS withgalaxiesradius(real and mock)(Efstathiouin the zone1995).of avHenceoidancethe(seefluctuationtext). in galaxy counts in a slab R and height along SGZ of 20 h21 Mpc, and the mean density aligned with the standard de Vaucouleurs' SGP is no more than max for selection effects, as a function of SGZ, averaged over SGX and21 ,1.5s perturbation on scale of ,40 h Mpc. This implies that knl is calculated within a sphere of radius RmaxSGY(including, within spheresthe of ever-larger radii, as indicated in each panel. 21 the SGP is only a modest perturbation on these large scales. slab). This overdensity reaches a maximum at RmaxThe SGP40 h isMpcseen as an enhancement at SGZ 0 in these plots to 21 3 (for which the volume of the slab is ,(46 h Mpc) ), dsgp 40 21 roughly R max 60 h Mpc, but is not apparent on larger scales. 0:48 ^ 0:06 and 0:46 ^ 0:09 for ORS and IRAS, respectively. At 3.3 The centre of the overdensity R 60 h21 Mpc; d 60 0:36 ^ 0:06 and 0:26 ^ 0:08 for max sgp ORS and IRAS, respectively. For reference, the rms3.2fluctuationsThe levelinof overdensityIn order to quantify the local overdensity, we needFigureto calculate4. Histogramsthe of the IRAS density field (corrected for selection effects) as a function of SGZ, averaged over SGX and SGY, within spheres q 2000 RAS, MNRAS 312, 166±176 Our visual impression is that there is indeed a flattened structure of radius as indicated in each panel. The error bars are Poissonian. in the galaxy distribution, aligned along SGZ 0, extending to 21 3 appreciable redshifts. To quantify this we calculate the overdensity cubes of volume (40 h Mpc) are drms , 0:37 and ,0.30 for optical (Stromlo/APM) and IRAS (1.2-Jy) surveys, respectively dsgp ; nsgp=knl 2 1 as a function of Rmax, where nsgp is evaluated within a circular cylinder centred on the Local Group with radius (Efstathiou 1995). Hence the fluctuation in galaxy counts in a slab 21 aligned with the standard de Vaucouleurs' SGP is no more than Rmax and height along SGZ of 20 h Mpc, and the mean density ,1.5s perturbation on scale of ,40 h21 Mpc. This implies that knl is calculated within a sphere of radius Rmax (including the 21 the SGP is only a modest perturbation on these large scales. slab). This overdensity reaches a maximum at Rmax 40 h Mpc 21 3 (for which the volume of the slab is ,(46 h Mpc) ), d 40 sgp 0:48 ^ 0:06 and 0:46 ^ 0:09 for ORS and IRAS, respectively. At 21 3.3 The centre of the overdensity Rmax 60 h Mpc; dsgp 60 0:36 ^ 0:06 and 0:26 ^ 0:08 for ORS and IRAS, respectively .For reference, the rms fluctuations in In order to quantify the local overdensity, we need to calculate the

q 2000 RAS, MNRAS 312, 166±176 Pie diagrams: structure of superclusters

Figure 7. The pieplot represents the radial distribution of galaxies along the projected rectangular strip shown in the lower panel. The strip covers a region 120°× 10°, orientated to lie along the filament. From the Norma cluster, lying 86° along the strip, the Norma supercluster is clearly seen as a wall of galaxies extending through the Pavo II cluster (at 71°) towards a point 20° along the strip. The Centaurus Wall appears as a smaller connection of galaxies, running almost parallel to the Norma supercluster at 2600 km s1. The void lying between the Norma supercluster and the Centaurus Wall is an extension of the massive Microscopium Void. Large scale distribution of galaxies and velocities Landscape of the MW neighborhood: 80Mpc Constrained simulations

Position of GA, Coma, Virgo, and Perseus-Pices. Our galaxy is a white circle at (0,0)

•Long waves (amplitudes and phases) are taken from observations. •Small scale- perturbations and dynamics are for LCDM cosmological model •N-body simulations at z=0 1G particles Dynamical Coma cluster range 1e5

Great Attractor

Constrained MW simulations

160 Mpc Perseus cluster Klypin, Hoffman, Gottlober Local Supercluster:

• Distribution of light and mass • Peculiar velocities (deviations from the Hubble ﬂow) • Voids

Galaxies in the Local Volume: ! - distances less than 10 Mpc Tikhonov& Klypin 2008 - 550 galaxies: Complete to M=-12 - Count voids: regions without galaxies

Data: Karachentsev & Co

21 Overabundance of Tikhonov & Klypin 2008 dwarf halos

Observations

Theory

22 Peculiar velocities of local galaxies: Voids:

• Distribution of light and mass in giant voids Voids Patiri et al 2006: SDSS R>10Mpc

25 Number-density Profiles of Voids: ! n(R)/

Patiri et al 2006

SDSS

LCDM

26 Large Scale Structure and Galaxies:

• Morphology - density relation • Effects of environment LSS: 300Mpc Ben Moore: PKDGRAV

Morphology -Density relation:

Bamford et al. Fraction of early-type galaxies as the function of density Gas stripping in Virgo cluster spiral galaxies

Spiral galaxies lose their gas and gradually get transformed into S0’s

Chung et al 2008 The main effect in morphology- density relation is that galaxies in denser environments are more massive and more massive galaxies tend to be ellipticals and red regardless of their environment. Fraction of red and early- type galaxies in low- density environment

There are many ellipticals in the ﬁeld where environment does not play a role

incompleteness Stellar mass Fraction of ellipticals for different colors and stellar masses

Small galaxies: • in low density environments are preferentially blue spirals • in high density they are red ellipticals Large galaxies: • preferentially red regardless of environment • in high density they are red ellipticals