Available via license: CC BY 4.0
Content may be subject to copyright.
Virgo Filaments. II. Catalog and First Results on the Effect of Filaments on Galaxy
Properties
Gianluca Castignani
1,2,3
, Benedetta Vulcani
4
, Rose A. Finn
5
, Francoise Combes
6
, Pascale Jablonka
3,7
,
Gregory Rudnick
8
, Dennis Zaritsky
9
, Kelly Whalen
10
, Kim Conger
8
, Gabriella De Lucia
11
, Vandana Desai
12
,
Rebecca A. Koopmann
13
, John Moustakas
5
, Dara J. Norman
14
, and Mindy Townsend
8
1
Dipartimento di Fisica e Astronomia, Alma Mater Studiorum Università di Bologna, Via Gobetti 93/2, I-40129 Bologna, Italy; gianluca.castignani@unibo.it
2
INAF—Osservatorio di Astrofisica e Scienza dello Spazio di Bologna, via Gobetti 93/3, I-40129, Bologna, Italy
3
Institute of Physics, Laboratory of Astrophysics, Ecole Polytechnique Fédérale de Lausanne (EPFL), Observatoire de Sauverny, CH-1290 Versoix, Switzerland
4
INAF—Osservatorio astronomico di Padova, Vicolo Osservatorio 5, I-35122 Padova, Italy
5
Department of Physics and Astronomy, Siena College, 515 Loudon Road, Loudonville, NY 12211, USA
6
Observatoire de Paris, LERMA, Collège de France, CNRS, PSL University, Sorbonne University, F-75014, Paris, France
7
GEPI, Observatoire de Paris, Université PSL, CNRS, Place Jules Janssen, F-92190 Meudon, France
8
University of Kansas, Department of Physics and Astronomy, 1251 Wescoe Hall Drive, Room 1082, Lawrence, KS 66049, USA
9
Steward Observatory, University of Arizona, 933 North Cherry Avenue, Tucson, AZ 85721-0065, USA
10
Department of Physics & Astronomy, Dartmouth College, 6127 Wilder Laboratory, Hanover, NH 03755, USA
11
INAF—Astronomical Observatory of Trieste, via G.B. Tiepolo 11, I-34143 Trieste, Italy
12
Spitzer Science Center, California Institute of Technology, MS 220-6, Pasadena, CA 91125, USA
13
Department of Physics & Astronomy, Union College, Schenectady, NY, 12308, USA
14
National Optical Astronomy Observatory, 950 North Cherry Avenue, Tucson, AZ 85750, USA
Received 2021 June 15; revised 2021 October 8; accepted 2021 October 20; published 2022 March 23
Abstract
Virgo is the nearest galaxy cluster; it is thus ideal for studies of galaxy evolution in dense environments in the local
universe. It is embedded in a complex filamentary network of galaxies and groups, which represents the skeleton of
the large-scale Laniakea supercluster. Here we assemble a comprehensive catalog of galaxies extending up to ∼12
virial radii in projection from Virgo to revisit the cosmic-web structure around it. This work is the foundation of a
series of papers that will investigate the multiwavelength properties of galaxies in the cosmic web around Virgo.
We match spectroscopically confirmed sources from several databases and surveys including HyperLeda, NASA
Sloan Atlas, NASA/IPAC Extragalactic Database, and ALFALFA. The sample consists of ∼7000 galaxies. By
exploiting a tomographic approach, we identify 13 filaments, spanning several megaparsecs in length. Long
>17 h
–1
Mpc filaments, tend to be thin (<1h
–1
Mpc in radius)and with a low-density contrast (<5), while shorter
filaments show a larger scatter in their structural properties. Overall, we find that filaments are a transitioning
environment between the field and cluster in terms of local densities, galaxy morphologies, and fraction of barred
galaxies. Denser filaments have a higher fraction of early-type galaxies, suggesting that the morphology–density
relation is already in place in the filaments, before galaxies fall into the cluster itself. We release the full catalog of
galaxies around Virgo and their associated properties.
Unified Astronomy Thesaurus concepts: Galaxy clusters (584);Virgo Cluster (1772);Large-scale structure of the
universe (902);Astronomy databases (83);Catalogs (205);Surveys (1671)
Supporting material: interactive figure, machine-readable tables
1. Introduction
Galaxies in the universe are not distributed uniformly at the
megaparsec scales. Large galaxy redshift surveys have revealed
that the universe has a prominent weblike structure made by
dense clusters and groups, elongated filaments, planar sheets,
and voids, called the cosmic web (Tifft & Gregory 1976; Joeveer
et al. 1978;Bondetal.1995). Galaxies are continuously
funneled into higher-density cluster environments through
filaments, which host ∼40% of the galaxies (e.g., Jasche et al.
2010;Tempeletal.2014;Cautunetal.2014). Therefore, the
analysis of filamentary structures can carry insights into the
assembly history of large-scale structures.
Characterizing the cosmic web and flow of galaxies in the
nearby universe is not an easy task, and many strategies have
been proposed, based on either observations (Tully et al.
2013,2016)or simulations (e.g., Libeskind et al. 2018,2020).
These methods often rely on the study of the geometry of the
galaxy density field or of the tidal field to reconstruct the
cosmic web, which indeed consists of a set of structures that are
anisotropic in shape (e.g., elongated filaments), multiscale
(groups, clusters, and filaments that can extend from a few to
100 Mpc), and are intricately connected (see, e.g., Cautun et al.
2014). The absence of both a common definition for the cosmic
filaments and a unique operative procedure to identify the
filamentary structures, as well as the lack of fields observed
with a very high sampling rate, have been major obstacles in
investigating not only the structure of the cosmic web but also
its impact on galaxy evolution.
Despite difficulties, filamentary structures of the cosmic web
have been identified in both simulations (e.g., Aragon-Calvo
et al. 2008; Cautun et al. 2014; Chen et al. 2015; Laigle et al.
2018; Kraljic et al. 2019; Kuchner et al. 2020,2021; Rost et al.
2021)and galaxy surveys (e.g., Tempel et al. 2014; Alpaslan
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April https://doi.org/10.3847/1538-4365/ac45f7
© 2022. The Author(s). Published by the American Astronomical Society.
Original content from this work may be used under the terms
of the Creative Commons Attribution 4.0 licence. Any further
distribution of this work must maintain attribution to the author(s)and the title
of the work, journal citation and DOI.
1
et al. 2014; Chen et al. 2016; Laigle et al. 2018; Kraljic et al.
2018; Malavasi et al. 2017,2020a,2020b). Many works have
also suggested that filaments affect the evolution of the
integrated properties of galaxies (e.g., Geach et al. 2011;
Koyama et al. 2011; Sobral et al. 2011; Mahajan et al. 2012;
Pintos-Castro et al. 2013; Tempel & Libeskind 2013a; Tempel
et al. 2013; Zhang et al. 2013; Koyama et al. 2014; Santos et al.
2014; Malavasi et al. 2017; Mahajan et al. 2018)and the
distribution of satellites around galaxies (Guo et al. 2014),at
any redshift, but results are still controversial. Overall, filament
galaxies tend to be more massive, redder, more gas poor, and
have earlier morphologies than galaxies in voids (Rojas et al.
2004; Hoyle et al. 2005; Kreckel et al. 2011; Beygu et al. 2017;
Kuutma et al. 2017). Some studies have also reported an
increased fraction of star-forming galaxies (Porter & Ray-
chaudhury 2006; Fadda et al. 2008; Porter et al. 2008; Biviano
et al. 2011; Mahajan et al. 2012; Darvish et al. 2014)and
higher metallicities and lower electron densities (Darvish et al.
2015)in filaments with respect to field environments.
Other studies even found evidence of a distinct impact of
filaments on galaxy properties and different gas phases.
Vulcani et al. (2019)showed that ionized Hαclouds in some
filament galaxies extend far beyond what is seen for other
noncluster galaxies. The authors suggest this may be due to the
effective cooling of the dense star-forming regions in filament
galaxies, which ultimately increases the spatial extent of the
Hαemission. Even atomic H Igas reservoirs are impacted by
the filament environments (Kleiner et al. 2017; Odekon et al.
2018; Blue Bird et al. 2020; Lee et al. 2021). The global
properties of galaxies’gas reservoirs as a function of distance
to the filament and local density are still debated. Some studies
claimed that galaxy and halo properties (e.g., luminosities,
masses, accretion rate, concentration)depend mostly on local
density, while the filament environment has no additional
effects beyond the ones related to the local density enhance-
ment (Yan et al. 2013; Eardley et al. 2015; Brouwer et al. 2016;
Goh et al. 2018).
Further investigations are therefore clearly needed. Our
approach is to focus on the area around Virgo, the benchmark
cluster in the local universe. It is embedded in a complex
filamentary network as it indeed belongs to the Laniakea
supercluster (Tully et al. 2014). The closeness of Virgo and its
associated high spectroscopic completeness makes its field
ideal for studies of galaxy evolution over a large range in
environments.
Numerous studies have characterized the galaxy population
of the Virgo cluster (Kim et al. 2014)and evaluated the
associated atomic and molecular gas content (Giovanelli et al.
2005; Chung et al. 2009; Boselli et al. 2014a,2014b,2014c),
dust (Davies et al. 2010)stellar masses (Ferrarese et al. 2012),
and star formation (Boselli et al. 2014d). However, galaxies in
the surrounding regions have received relatively little attention.
Tully (1982)identified prolate and oblate overdensities of
galaxies connected to the cluster. Nonetheless, due to the
limited size of their sample, these elongated structures were not
clearly revealed as conventional narrow filaments. A better
characterization of these structures requires improved statistics
from larger galaxy samples, particularly those with fainter
galaxies. Building upon Tully’s results, Kim et al. (2016)used
the seventh release of the Sloan Digital Sky Survey (SDSS;
Abazajian 2008)combined with the HyperLeda catalog
(Makarov et al. 2014)to more firmly identify the filamentary
structures within an extensive volume around the Virgo cluster.
While providing a detailed characterization of the filaments
around the Virgo cluster, Kim et al. (2016)did not release their
environmental classification.
In Castignani et al. (2022, from now on Paper I)we therefore
assemble an independent catalog of Virgo and the surrounding
volume. We accomplish this by matching and vetting several
existing catalogs, with the intent of releasing a comprehensive
catalog of galaxies in the filaments around Virgo, extending out
to ∼12 virial radii in projection (i.e., ∼24 Mpc)from the
cluster, with a small fraction (7%)of galaxies reaching even
higher distances, up ∼40 Mpc from Virgo. Two strengths of
this catalog are that (1)we have high completeness because we
merge sources from multiple catalogs of local galaxies, and (2)
we have low contamination because we visually inspect every
source in our sample.
Here we describe more in detail our adopted procedure and
release the catalog (see the Appendix). With respect to Paper I,
we refine both the source catalog by visually inspecting each
object to remove duplicates, stars, and “shredded”galaxies and
by excluding galaxies in the southern hemisphere where SDSS
has poor spectroscopic coverage and filament definition (see
Section 3.3).
Overall, we consider a larger survey area in the northern
hemisphere than that covered by Kim et al. (2016). This allows
us to identify and characterize additional filamentary structures
to the north and east of the Virgo cluster that were not
identified by Kim et al. (2016).
This catalog forms the foundation of a series of papers aimed
at investigating the effect of the filament environment on
processing the gas of galaxies and on global properties such as
star formation and stellar content. The first exploration of the
catalog has been presented in Paper Iwhere we analyzed
spatially integrated CO and H Iobservations for a subset of
filament galaxies. We found a clear progression as one moves
from field to filament and cluster in that galaxies in denser
environments have a lower star formation rate, a higher fraction
of galaxies in the quenching phase, an increasing proportion of
early-type galaxies, and decreasing gas content. In addition,
galaxies in the densest regions in filaments tend to be deficient
in their molecular gas reservoirs, which fuel star formation.
These results suggested that processes that lead to star
formation quenching are already at play in filaments. Following
this study, we are carrying out follow-ups at different
wavelengths, with the aim of linking the galaxy stellar
properties to the galaxy gas content. In particular, we will
investigate the physical mechanisms responsible for the
preprocessing using ongoing high-resolution observations in
both CO and H I. In parallel, for a few hundred filament
galaxies, we are conducting a Hαimaging survey to map the
spatial distribution of the hot gas and to derive integrated star
formation rates. All of these campaigns will be described in
forthcoming papers.
The outline of this paper is the following. In Section 2we
describe how we build the catalog of galaxies around Virgo. In
Sections 3and 4we characterize the cosmic-web environment
around Virgo. In Section 5we contrast the different
parameterizations of the environment and in Section 6we
investigate the interplay between galaxy properties and their
cosmic-web environment and describe our results. In Section 7
we draw our conclusions and summarize the paper.
2
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
Throughout this paper, we assume a Hubble constant of
H
0
=100 hkm
−1
Mpc
−1
, where h=0.74 (e.g., Tully et al.
2008; Riess et al. 2019). Magnitudes are reported in the AB
system.
2. The Spectroscopic Parent Catalog
To assemble a spectroscopic sample of galaxies around
Virgo (R. A. =187°. 70, decl. =12°. 34, J2000), we start by
creating a catalog from the union of HyperLeda (Makarov et al.
2014),
15
the NASA Sloan Atlas
16
(NSA; Blanton et al. 2011),
and the ALFALFA α100 sample (Haynes et al. 2018)in the
region covered by 100°<R.A. <280°,−1°.3 <decl. <75°,
and recession velocities 500 <v
r
<3300 km s
−1
. The southern
limit coincides with the southern limit of the SDSS spectro-
scopic survey. We adopt this cut because we want high
spectroscopic sampling to robustly identify and characterize
filaments. However, this choice is different from what was
done by Kim et al. (2016), who also characterized Virgo
filaments to the south. The lower velocity cut is dictated by the
need to avoid stars and galactic contamination. The higher
velocity cut is set by the need to include all filaments, which are
mostly located farther than Virgo (cz ∼1000 km s
−1
; Mei et al.
2007).
To build the sample, we start with all sources from
HyperLeda that are classified as galaxies. We then match the
HyperLeda sources to version 1 of the NSA, using a search
radius of 10″and a maximum velocity offset of 300 km s
−1
.
This updated version of the NSA extends to larger distances
and contains additional fitted parameters
17
relative to version 0
presented in Blanton et al. (2011).
We initially allow for the same NSA source to be matched to
multiple HyperLeda sources, and we later eliminate these
duplicates by visual inspection (see below). We then append as
new catalog entries any additional NSA sources that were not
matched to HyperLeda. We repeat a similar match to version 0
of the NSA (Blanton et al. 2011)because some of the sources
and redshifts differ between the two versions of the NSA
catalogs. Versions 0 and 1 of NSA are complementary in terms
of the number of galaxies that fall in the region of interest,
which motivates the choice of querying both catalogs.
We then match the list of HyperLeda+NSA galaxies to the
ALFALFA α100 sample (Haynes et al. 2018), limited to
ALFALFA galaxies with 500 <v
r
<3300 km s
−1
. The north-
ern limit of the ALFALFA survey is decl. =36°, so we do not
have ALFALFA coverage for our full survey area. However,
54% of our decl. <36°sources are matched to an ALFALFA
source, and this provides a rich sample for future studies on
how the atomic gas reservoir is affected by the filament
environments. As a blind H Isurvey, ALFALFA detects a
higher fraction of low-mass star-forming galaxies relative to
optical surveys (e.g., Durbala et al. 2020). However, at the
relatively close distance of Virgo, we find only nine ALFALFA
sources that are not already in either the Hyperleda or NSA
catalogs. We add these nine sources to our catalog.
We assign a position (R.A., decl.)to each galaxy based on
information in the source catalogs. We assign HyperLeda
coordinates if they are available. If HyperLeda is not available,
we then use NSA version 0, followed by NSA version 1, and
ALFALFA. We assign recession velocities by the same
process.
Next, we add to the sample 110 galaxies that have redshift-
independent distances in the NASA/IPAC Extragalactic
Database compendium of distances based on primary and
secondary indicators (NED-D; Steer et al. 2017). These 110
galaxies have redshift-independent distances that correspond to
cosmological velocities in the range of 500–3300 km s
−1
, but
they are missing in our catalog as their observed recession
velocities are less than <500 km s
−1
. Some of these sources are
Virgo cluster members that are located near the caustics and
thus have the largest deviation in velocity with respect to that
of Virgo.
Finally, to compile as clean a sample as possible in the area
of interest, we visually review each galaxy in our catalog to
remove shredded galaxies, duplicates, and spurious objects. We
also flag galaxies with nearby stars that might affect the
photometry, and we recenter the coordinates of some galaxies,
as needed.
To identify any remaining stars, we cross-match with the star
catalog used by the Legacy Survey.
18
This catalog is built from
Tycho-2 (MAG
VT
<13)and Gaia-DR2 sources (G<16).We
look for matches within r<10″of our sources, and we find an
additional two stars, which we remove.
While we require all of the sources to have a galaxy
classification, we find that a number of HyperLeda sources are
instead globular clusters in nearby galaxies, as identified by Ko
et al. (2017). We therefore remove all sources with the prefix
“S”in the Ko et al. (2017)catalog. On the basis of our cleaning
procedure mainly aimed at removing duplicates and shredded
objects, we found that ∼4% of HyperLeda sources in our
region of interest are misclassified as galaxies.
For each galaxy, we also query the NED server to get its
official NED name. We use the object name from HyperLeda
as input if it is available. If not, we then use the NSA name and
the ALFALFA/AGC name. If NED does not return a match by
name for any of the catalog names, we then match the source
by position, using a search radius of 10″. We include the input
name used in the NED search as well as the official NED name
in our table. Note that for some galaxies, we are not able to find
a corresponding NED name.
Our final sample contains 6780 galaxies. The contributions
from the different input catalogs are broken down in Table 1.
The NSA v1 (v0)catalog provides 157 (122)galaxies that are
not in the v0 (v1)version of the catalog. We stress that only
nine galaxies are in the ALFALFA α100 sample, but not in the
union of the HyperLeda and NSA source samples.
Table 1
Statistics of the Parent Sample
Catalog No. of Galaxies Fraction
Final 6780 1
HL 6622 0.98
NSA v1 5280 0.78
NSA v0 5245 0.77
α100 2336 0.34
NED-D 1959 0.29
15
http://leda.univ-lyon1.fr/
16
http://nsatlas.org
17
https://www.sdss.org/dr13/manga/manga-target-selection/nsa/
18
https://portal.nersc.gov/cfs/cosmo/data/legacysurvey/dr9/masking/
gaia-mask-dr9.fits.gz
3
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
2.1. Photometry
We cross-match our catalog to the ninth public data release
of the DESI Legacy Imaging Surveys (DR9; Dey et al. 2019),
using a search radius of 10″. The Legacy Survey covers 14,000
deg
2
of extragalactic sky visible from the northern hemisphere
in three optical bands (g, r, z)and four infrared bands. In this
paper, we utilize only the r-band photometry from the DR9
catalogs to apply a magnitude cut and analyze galaxy
properties in an absolute-magnitude-complete sample.
The available DR9 photometric catalogs are based on the
Tractor fitting (Lang et al. 2016); like all automated photometry
codes, Tractor struggles with providing meaningful models to
clumpy, well-resolved galaxies. We therefore have efforts
underway to measure custom photometry from the Legacy
imaging that is optimized for large, nearby galaxies. In a
forthcoming paper, we will present the multiband photometry
for our entire catalog of sources in the field of Virgo, with a
careful treatment of the extended galaxies (>05 in size).
As most of the spectroscopic redshifts for galaxies in the
catalog come from the SDSS, we adopt the SDSS completeness
limit of r=17.77. This corresponds to an absolute limit of
M
r
=−15.7 at a distance modulus of 33.5, approximately the
upper limit of the survey.
2.2. The Final Catalog
To summarize, we have assembled a catalog of galaxies with
500 <v
r
<3300 km s
−1
located in the region surrounding the
Virgo cluster (up to ∼12 virial radii, i.e., 24 Mpc, in projection
from the center of Virgo)by combining the sources present in
HyperLeda, NSA (v0 and v1), ALFALFA, and NED-D. This
catalog is cleaned from spurious sources, stars, and duplicates
and represents a unique starting point to define the cosmic web
around the Virgo cluster, as detailed in what follows.
The final catalog (Table 1)contains 6780 galaxies, 3528 of
which are above the absolute magnitude limit M
r
=−15.7
(+M
3
r, Blanton et al. 2005). The subsample of galaxies
with M
r
<−15.7 corresponds to a volume-limited sample, with
the M
r
limit corresponding to the SDSS m
r
=17.77 spectro-
scopic completeness limit at the maximum distance of the
galaxies in our catalog. As we will describe in Section 3.3,we
define different volume-limited subsamples appropriate for the
distance range of each filament.
Figure 1shows the projected spatial distribution of the final
sample, color-coded by recession velocity (left panel), and the
distribution of the recession velocity (right panel)for both the
entire sample and subsample above the absolute magnitude
limit.
Hereafter, to identify galaxies in different global environ-
ments (Sections 3and 4.1), we will make use of the full catalog
of ∼7000 galaxies. When we compute local densities to
characterize the properties of galaxies in the different
environments (from Section 4.2 onward), we will adopt the
magnitude-complete sample.
3. The Virgo Cluster and Its Infalling Filaments
In this section we provide a characterization of the cosmic
web around Virgo using the catalog of galaxies assembled
above. We will rely on a widely used description of the cosmic
flow around Virgo (Mould et al. 2000)and on redshift-
independent distances, when available (Steer et al. 2017).
To properly investigate the effect of the megaparsec-scale
environment on galaxy properties, it is necessary to provide
both local and more global parameterizations of the density as,
depending on the scale probed, different physical processes
might shape galaxy properties. For example, the frequency of
galaxy–galaxy interactions depends on the local density of
galaxies, whereas gas accretion onto galaxies varies depending
on whether the galaxy is a central or satellite galaxy in the
parent halo mass. After computing the distances for all galaxies
(Section 3.1), we will thus assign to all galaxies a global
environment depending on whether they are in the Virgo
cluster (Section 3.2)or in filamentary structures (Section 3.3).
In Section 4.1 we will further investigate the presence of
groups within filaments and assemble a sample of pure field
galaxies, aided by the Kourkchi & Tully 2017 group catalog.
We will finally evaluate local densities for all galaxies
(Section 4.2), regardless of all of the above memberships.
Figure 1. Left: spatial distribution of galaxies around the Virgo galaxy cluster, up to ∼12 virial radii from its center, in projection. Points show galaxies coded
according to their recession velocities. The red rectangle shows the region of the Virgo cluster, which is examined in more detail in Figure 3. Right: radial velocity
distribution of all galaxies in our catalog. The black solid line shows the distribution of galaxies brighter than the absolute magnitude completeness limit of the catalog
(M
r
=−15.7).
4
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
We will then use these characterizations to (i)determine the
filament profiles (Section 3.3.2),(ii)compare the different
definitions of environment (Section 5), and (iii)describe the
dependence of galaxy properties on the different environments
(Section 6).
3.1. Cosmic Distances
To characterize the positions of all galaxies around Virgo,
we convert heliocentric velocities v
r
to intrinsic distances of the
sources, according to the following steps.
First, we match our sample using the NED name to the
NED-D catalog (Steer et al. 2017). The search yields a match
for 1959 sources—corresponding to 29% of the total sample—
and for these sources in what follows we will adopt the Steer
et al. (2017)distances (D
z−independent
)as the final cosmic
distances. We calibrate all D
z−independent
assuming
H
0
=74 km s
−1
Mpc
−1
used in this work. In the cases where
the Steer et al. (2017)catalog provides multiple estimates for a
given source, we adopt the median distance.
Second, we compute intrinsic distances following Mould
et al. (2000), using their method for correcting observed
recession velocities for peculiar motions associated with
various attractors in the local universe. We derive the
correction v
LG
of the observed heliocentric velocity of our
galaxies to the centroid of the Local Group (LG),asin
Equation (A1)from Mould et al. (2000). Then we estimate the
correction v
in,Virgo
that takes into account the infall toward the
Virgo attractor as in Equation (1)by Mould et al. (2000).
Distances and radial velocities relative to the Virgo center are
calculated by means of the cosine theorem (e.g., Karachentsev
& Nasonova 2010). A cosmic velocity of ∼1016 km s
−1
is
assumed for Virgo, as found in NED. It is obtained by
correcting Virgo heliocentric velocity to the LG centroid for
our infall velocity and for the infall of Virgo into the Great
Attractor, as described in Appendix A of Mould et al. (2000).
We also assume a Virgo density profile ρ(r)∝r
−2
and an
amplitude v
fid
=200 km s
−1
for the Virgo infall velocity
(Mould et al. 2000). Model-corrected velocities v
model
are then
derived as follows:
()=+ +vvvv.1
rmodel LG in,Virgo
Here we ignore higher-order corrections in Equation (A2)by
Mould et al. (2000)that are due to the infall of our galaxies
toward the Great Attractor and Shapley supercluster. We also
assume a linear dependence between velocities and distances.
Figure 2shows the comparison between the model-corrected
distances (D
model
)and D
z−independent
for galaxies with both
redshift-independent and model-corrected distances. The med-
ian logarithmic difference is ()
=
-
DDlog zmodel independent
-
-
+
0.03 0.18
0.12. Here the reported uncertainties correspond to the
1σconfidence interval. The comparison yields a negligible bias
and an rms scatter of ∼0.1 dex, which is consistent with that
found in recent studies of the local universe (Leroy et al. 2019).
The small differences, well within the uncertainties, between
these values and those reported in Paper Iare due to the
different southern limits adopted in the two works, as here we
do not consider galaxies at negative declinations as Paper Idid.
Although we find an overall agreement between redshift-
independent and model-corrected distances, we do see an
increased dispersion in the data points in Figure 2along the y-
axis at ∼17 Mpc, which corresponds to the distance of Virgo
(Mei et al. 2007). The model correction for Virgo galaxies is
more uncertain because peculiar velocities become significant
as we approach the Virgo cluster.
Overall, from the comparison presented in Figure 2we
conclude that for most of the galaxies in our sample with no
redshift-independent distance, the model-corrected version is
reliable enough to determine galaxy 3D positions and local
density estimates (see Section 4.2). The model-corrected
distances might not be as reliable for the Virgo cluster
members, and in principle, this could impact our estimates of
local density. Nonetheless, we will show in Section 6that these
uncertainties will not significantly affect our results.
To summarize, our adopted cosmic distances and velocities,
D
cosmic
and v
cosmic
, are the redshift-independent distances and
velocities, when available, and those derived as in Equation (1)
for the remaining sources.
We then make use of the super-Galactic (SG)coordinate
system, which was developed by Gérard de Vaucouleurs. This
coordinate frame has the equator aligned with the SG plane,
which consists of a planar distribution of nearby galaxy
clusters. The SG system is thus ideal for studies of the cosmic
web in the local universe. Therefore, assuming a linear
relationship between v
cosmic
and distance, galaxies have been
mapped into the Cartesian SG frame. In this frame galaxy
positions are defined in terms of their SG coordinates SGX,
SGY, and SGZ (Tully et al. 2008). We note that the Virgo
cluster center has (SGX; SGY; SGZ)=(−2.26; 9.90; −0.42)
h
−1
Mpc in the SG coordinate frame. At the coordinates of
Virgo the SGY direction approximately corresponds to the line
of sight.
3.2. Membership of the Virgo Cluster
To identify galaxies belonging to the Virgo cluster, we select
galaxies within 3.6 h
−1
Mpc from the Virgo cluster center in
the 3D SG coordinate frame. The chosen radius corresponds
approximately to ∼3r
200
, with r
200
=1.09 h
−1
Mpc
Figure 2. Comparison between the redshift-independent distances obtained
from Steer et al. (2017)and those inferred from v
model
for the galaxies present
in the Steer et al. (2017)catalog (black points). The green line is the 1:1
relation. The red line shows the linear fit to the point, while the shaded red
region denotes the corresponding ±1σscatter.
5
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
(McLaughlin 1999)the radius that encloses 200 times the
critical matter density. The position of the 311 Virgo members
selected in this way in the phase-space diagram is shown in
Figure 3. Overall, they fall within the region delimited by the
caustics that are defined following the prescription by Jaffé
et al. (2015), assuming the r
200
radius and a concentration
parameter of 2.8 as reported by McLaughlin (1999). As the
adopted definition is rather conservative, we also consider as
cluster members those galaxies that fall within the cluster
region delimited in the phase-space diagram by the caustics,
regardless of their position in the SG coordinates. The final
cluster member sample is the union of the members defined in
SG coordinates and those defined using the phase space, for a
total of 1152 galaxies (526 above the magnitude completeness
limit).
3.3. Filamentary Structures
Moving beyond the cluster, we aim to characterize its
surrounding cosmic web in 3D. We start by considering the
eight filamentary structures presented in Tully (1982), Kim et al.
(2016):theW–M Sheet located to the south of Virgo; the nearby
Ursa Major cloud in the North; the Virgo III filament to the south
of Virgo; the extended NGC 5353/4filament, with the
corresponding group at the end of it; the Canes Venatici filament
just north of NGC 5353/4filament; and the Leo II A, Leo II B,
and Leo Minor filaments belonging to the Leo cloud to the
northwest. To test the reliability of these filaments, we construct a
series of different (SGX; SGY; SGZ)volume slices with an
arbitrary depth of 4 h
−1
Mpc along the SGY axis. Selected
structures are confirmed by visual inspection of the (SGX; SGZ)
projection of each slice, looking for overdense and long (i.e.,
filamentary)galaxy distributions. All candidate structures are
present in consecutive slices. During this visual inspection of the
distribution of galaxies in the (SGX; SGZ)plane, we identify five
additional structures that were not reported in Kim et al. (2016)
and that will enter our final filament sample. We name these
respective structures the Leo Minor B, Bootes, Serpens, Draco,
and Coma Berenices filaments, where the names of these
filaments derive from the dominant constellation that they are in.
As further discussed in Section 3.3.1, all these structures have a
least one main counterpart in the V8k catalog of nearby sources
and structures (Courtois et al. 2013).
Nevertheless, we stress that the goal of this work is not to
provide a complete census of all filaments around Virgo.
Instead, we provide a detailed characterization of the filaments
that have the highest density contrast relative to the surround-
ing field as determined by visual inspection, including those
already known in the northern hemisphere.
The 13 structures all fall within the cuboid enclosed by the
following limits:
()
()
()
-< <
<<
-< <
-
-
-
h
h
h
13 SGX Mpc 20,
2SGY Mpc 38,
15 SGZ Mpc 33.
1
1
1
These limits correspond to a more extended region in the
northern hemisphere than that considered by Kim et al. (2016),
which allows us to have a more comprehensive characterization
of the large-scale structures around Virgo than previous studies.
Note that the (SGX; SGZ)coordinate frame approximately
corresponds to the plane of the sky where filamentary structures
are better defined, while the SGY axis is associated with the
line of sight, and thus more impacted by positional errors
arising from distance uncertainties.
Similar to Kim et al. (2016), for each filamentary structure we
consider an associated parallelepiped Ωin the 3D SG frame, large
enough to conservatively enclose all galaxies that belong to the
structure. We set the parallelepiped dimensions after the visual
inspection of the filamentary structure in (SGX; SGY; SGZ)
coordinates, with different projections. We then determine the
filament spines by fitting the locations of the galaxies in SG
coordinates. We parameterize the spine of each filament by fitting
a third-order polynomial curve γ:[0,1]→Ω,suchthat
()
g
=+++
abcdtttt
32 .Herea,b,c,and Î
d
3are
the curve parameters with their origin coincident with the Sun, as
this is the case for the SG (X, Y, Z)coordinate system. We then
Figure 3. Left: phase-space diagram for sources in the field of the Virgo cluster. The solid lines show the radial dependence of the escape velocity in the phase-space
diagram, according to the prescription by Jaffé et al. (2015). Cluster members defined in the super-Galactic (X, Y, Z)coordinate frame (red points)or within the region
(blue points)delimited by the caustics (i.e., the two solid lines)are distinguished from the remaining sources in the field of Virgo cluster (gray points). Right: projected
distribution of the cluster members on the sky.
6
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
perform a fit by minimizing the sum of the distance squares of
each galaxy in Ωto the filament spine. In Table 2we report the
spatial extent of the filaments, which span a wide range in length,
between L∼(8−26)h
−1
Mpc. In Table 3we report the best-fit
parameters of our fits to the filament spines. In Table 4we provide
different points that sample the filament spines both in projection
and in the 3D SG frame. In the same table, we also report the
position angles of the tangent vectors, along each filament spine.
To identify filament members we select galaxies found within
2h
−1
Mpc of the spine, with the radial cut selected to minimize the
contamination from the field (Lee et al. 2021; Galárraga-Espinosa
et al. 2020).Weverified a posteriori that all considered
filamentary structures are overdense and elongated over several
megaparsecs in length. Indeed, as further outlined in Section 3.3.2,
the density contrast, evaluated as the ratio between the average
number density of galaxies within 1 h
−1
Mpc from the filament
spine relative to the field value, ranges between ∼3−18, which
thus strengthens the reliability of the selected filaments.
As an example of the outcome of our procedure, Figure 4
shows the selected parallelepiped in the SG frame within which
the Virgo III filament is embedded. The filament spine and
filament members within 2 h
−1
Mpc from the spine are
highlighted. The latter are color-coded by their local density
(see Section 4.2)to highlight density variations along the
filament. These variations are also due to the presence of
groups within the filament (see Section 4.1). The complexity of
these structures motivates further characterization of the
environment, even within filaments.
Figure 5shows the spatial distribution of the 2118 galaxies
belonging to the identified structures. Filaments are sorted by
increasing distance from us, and this color scheme is adopted
throughout the paper to help the reader track the different
filaments. It appears evident that different filaments exhibit
different properties in terms of their distance, richness, and
structure. In particular, the W–M Sheet has a planar morphology,
as will be further discussed in the following sections. The Leo
filaments were originally classified as a single cloud by Tully
(1982). Furthermore, the Ursa Major Cloud and the W–M Sheet
overlap with the Virgo cluster periphery. Indeed, 418 cluster
galaxies are also members of the Ursa Major Cloud (214)or the
Table 2
Detected Filaments and Spatial Extent of Their Spines
Structure SGX SGY SGZ R.A. Decl. L
(h
−1
Mpc)(h
−1
Mpc)(h
−1
Mpc)(deg)(deg)(h
−1
Mpc)
(1)(2)(3)(4)(5)(6)(7)
Leo Minor F 0.59 ∼5.64 4.06 ∼6.30 −2.83 ∼−1.77 120.27 ∼160.23 23.12 ∼52.30 7.63
Canes Venatici F 0.64 ∼3.97 5.89 ∼14.28 1.35 ∼4.79 197.46 ∼203.68 34.47 ∼44.13 10.53
Bootes F 5.9 ∼10.59 15.71 ∼22.83 5.91 ∼11.51 201.54 ∼216.01 43.31 ∼60.89 11.83
Ursa Major Cloud 0.50 ∼8.74 2.67 ∼13.95 0.12 ∼1.37 177.62 ∼186.07 34.38 ∼57.20 15.44
Leo II B F 2.42 ∼13.35 13.65 ∼13.97 −8.37 ∼−4.30 131.00 ∼163.99 27.82 ∼48.15 12.67
Leo II A F 0.43 ∼9.18 12.25 ∼14.47 −14.26 ∼−6.73 126.23 ∼156.97 15.79 ∼33.13 13.93
Virgo III F −10.94 ∼−5.51 11.93 ∼17.32 3.40 ∼11.09 207.31 ∼224.80 2.32 ∼5.42 11.72
Leo Minor B F 5.84 ∼10.83 18.28 ∼21.53 −7.01 ∼−5.41 152.99 ∼163.15 34.13 ∼41.94 7.95
W–M Sheet −9.28 ∼−3.41 20.30 ∼23.27 −2.67 ∼−1.91 183.30 ∼187.53 1.59 ∼15.15 8.45
NGC 5353/4F −12.57 ∼9.42 25.96 ∼27.67 0.30 ∼9.20 193.75 ∼204.04 2.09 ∼47.84 24.01
Serpens F −5.48 ∼−0.99 11.02 ∼17.01 9.47 ∼33.14 230.39 ∼256.6 10.82 ∼24.33 25.63
Draco F 13.98 ∼18.77 16.71 ∼21.50 14.90 ∼22.51 227.08 ∼259.51 58.76 ∼60.92 12.31
Coma Berenices F 2.15 ∼6.04 12.89 ∼38.31 −4.92 ∼−1.89 173.06 ∼175.40 30.13 ∼35.75 26.27
Note. Column description: (1)filament name; range in SG coordinates (2–4)and in projected space (5–6)spanned by the filament spine; (7)filament spine length.
Table 3
Best-fit Parameters for the Filament Spines
Structure abcd
(a
x
,a
y
,a
z
)(b
x
,b
y
,b
z
)(c
x
,c
y
,c
z
)(d
x
,d
y
,d
z
)
(h
−1
Mpc)(h
−1
Mpc)(h
−1
Mpc)(h
−1
Mpc)
Leo Minor F −2.00 –2.00 2.00 −5.00 −5.00 −4.32 11.59 6.52 1.38 0.59 4.54 −1.90
Canes Venatici F 2.00 −2.00 2.00 −0.53 −5.00 −1.37 1.87 15.38 2.80 0.64 5.89 1.35
Bootes F −2.00 1.08 −2.00 −5.00 −5.00 −5.00 2.62 11.03 12.28 10.28 15.71 5.91
Ursa Major Cloud 2.00 −2.00 1.52 5.00 −5.00 −1.32 1.24 18.28 1.05 0.50 2.67 0.12
Leo II B F 2.00 −2.00 2.00 3.62 1.90 5.00 5.31 0.09 −10.28 2.42 13.66 −4.30
Leo II A F −2.00 −2.00 −2.00 −5.00 −5.00 −5.00 15.74 7.39 −0.54 0.43 12.25 −6.73
Virgo III F 2.00 2.00 −2.00 −1.61 0.10 −5.00 −5.82 3.29 14.65 −5.51 11.93 3.40
Leo Minor B F 2.00 −1.27 2.00 5.00 −5.00 −4.14 −2.24 8.77 0.55 6.07 18.28 −5.42
W–M Sheet 2.00 2.00 −2.00 5.00 5.00 1.52 −1.20 −8.67 1.00 −9.21 23.27 −2.67
NGC 5353/4F −2.00 2.00 −2.00 −5.00 1.32 −4.48 28.99 −4.33 15.38 −12.57 27.67 0.30
Serpens F 2.00 −2.00 2.00 5.00 −5.00 5.00 −2.89 12.74 16.66 −5.10 11.02 9.47
Draco F −2.00 −2.00 −2.00 −5.00 −5.00 −5.00 11.25 2.50 14.56 13.98 21.21 14.90
Coma Berenices F −2.00 2.00 1.59 −5.00 3.46 −4.17 10.04 19.95 −0.45 2.15 12.89 −1.89
Note. Each spine is parameterized with a polynomial curve γ:[0,1]→Ω, such that γ(t)=at
3
+bt
2
+ct+d. The best-fit values of the parameters a,b,c, and dare
provided in (SGX; SGY; SGZ)coordinates for all filamentary structures considered in this work.
7
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
W–M Sheet (204). This is primarily due to the difficulty in
unambiguously distinguishing Virgo cluster members from those
of nearby correlated structures, as further discussed in previous
studies (e.g., Kim et al. 2014;Kourkchi&Tully2017).
Each filament is located at a different mean distance, and the
very conservative absolute magnitude limit for the catalog was
set by the most distant galaxies in the entire sample. We therefore
compute an absolute magnitude limit that is appropriate for each
filament (Figure 6), which might be useful for specificstudies
(e.g., comparing the local density of filament galaxies with the
density in the surrounding field, see Figure 14).Specifically, we
compute the completeness limit by computing the distance
modulus from the distance encompassing 95% of galaxies in any
given filament. Table 5reports the magnitude limit and the
number of galaxies above it for each filament.
3.3.1. Comparing Different Filament Determinations
As we follow the approach presented by Kim et al. (2016),we
now briefly compare our results with theirs. In Table 5we report
the number of filament members identified by Kim et al. (2016)
for the seven filaments in common. For these filaments the ratio of
the total number of members found in this work to that reported
by Kim et al. (2016)ranges between ∼0.6−2.3, with a median
value of 1.3. This wide range of values is due to both the different
input catalogs used (our catalog has been carefully cleaned of
duplicates and includes sources from additional surveys)and to
the different filament membership assignments.
Memberships to filaments around the Virgo cluster can be
retrieved also from the Tempel et al. (2014)catalog. However, our
approach has been fine-tuned to characterize filaments specifically
around Virgo, whereas Tempel et al. (2014)have searched for a
largesampleoffilaments in the Sloan Digital Sky Survey (SDSS)
over a wider field and up to larger distances (up to 450 h
−1
Mpc).
Their method approximates the filamentary network using a
random configuration of small segments (thin cylinders).Ifwecut
theTempeletal.(2014)catalog at our velocity limit (z<0.012),
we retain only 1281 galaxies and 774 of these are associated with
39 filamentary structures made ofmorethan10galaxies.This
includes filaments in the location of the Serpens, Bootes, Canes
Venatici, NGC 5353/4filaments, the Ursa Major Cloud, and the
W−M Sheet, but these filaments have many fewer members.
Overall their filaments are much less populated: the median
number of filament members found within 1 h
−1
Mpc from the
filament spine is 15. It appears therefore that to carefully
characterize filaments in the local universe, it is not appropriate
to apply a general approach that is optimized for a much larger
(z<0.155)redshift range. The above considerations motivated us
to exploit a method that is tailored to the case of the local universe
and specifically to filaments around Virgo.
In this context, it is worth mentioning the V8k catalog of
nearby sources and structures that is discussed by Courtois et al.
(2013)and is part of the Extragalactic Distance Database (Tully
et al. 2009).
19
This catalog provides a census of large-scale
structures in the local universe and their members. By cross-
matching, via membership, the structures considered in this
Table 4
A Sample of Filament Points
Filament ID
point
R.A. Decl. SGX SGY SGZ PA
(deg)(deg)(h
−1
Mpc)(h
−1
Mpc)(h
−1
Mpc)(deg)
Virgo III 1 207.312 5.419 −5.514 11.929 3.397 92
Virgo III 2 207.875 5.398 −5.572 11.961 3.543 92
Virgo III 3 208.426 5.374 −5.631 11.994 3.688 93
Virgo III 4 208.966 5.348 −5.69 12.027 3.831 93
Virgo III 5 209.494 5.32 −5.749 12.061 3.974 93
Virgo III 6 210.012 5.289 −5.809 12.094 4.116 94
Virgo III 7 210.518 5.257 −5.868 12.127 4.257 94
Virgo III 8 211.014 5.223 −5.928 12.16 4.397 94
Virgo III 9 211.499 5.187 −5.989 12.194 4.535 94
Virgo III 10 211.973 5.15 −6.049 12.227 4.673 95
Draco 1 227.076 58.758 13.983 21.212 14.895 67
Draco 2 227.476 58.843 14.095 21.237 15.04 68
Note. Column description: (1)filament name; (2)integer index associated with the curve parameter t;(3–4)projected coordinates, (5–7)SG coordinates, and (8)
position angle for the filament spine at the index reported in column (2).
(This table is available in its entirety in machine-readable form.)
Figure 4. Galaxies in the vicinity of the Virgo III filament. Galaxies with
in 2 h
−1
Mpc are color-coded by the 3D local density, and galaxies with
separations greater than 2 h
−1
Mpc are shown with the gray points. The
filament spine is shown with the black curve. Interactive 3D plots for all
filaments are available for download from Zenodo at doi:10.5281/
zenodo.6341838.
19
http://edd.ifa.hawaii.edu/index.html
8
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
Figure 5. Spatial distribution of galaxies around the Virgo galaxy cluster. Gray points show all galaxies, colored points show galaxies belonging to the different
filaments. This color scheme will be kept in all plots.
Figure 6. Absolute magnitude M
r
as a function of velocity V
model
(black line)
and magnitude limit of the survey (horizontal dashed line). Stars represent the
magnitude limit proper of each filament separately, and the green square the
magnitude limit of the cluster. These limits are obtained as the magnitude
including 95% of the data. For display purposes, points of the different
filaments are also shown in colors, with an arbitrary vertical shift to avoid the
superimposition of the points.
Table 5
Number of Galaxies in Each Filament
Structure N
gal
N
gal
@
M
rlim
M
rf,li
m
N
gal
@
M
rf,li
m
K16
Virgo Cluster 1152 526 −15.41 570
Leo Minor F 124 13 −12.86 62 54
Canes Venatici F 96 24 −14.01 48 51
Bootes F 169 113 −15.07 136
Ursa Major Cloud 580 117 −14.00 217
Leo II B F 63 28 −14.52 43 105
Leo II A F 145 53 −14.64 97 180
Virgo III F 206 115 −14.69 148 181
Leo Minor B F 39 28 −14.90 29
W–M Sheet 345 198 −14.96 250 256
NGC 5353/4 F 133 90 −15.34 106 102
Serpens F 65 34 −15.35 39
Draco F 48 44 −15.61 45
Coma Berenices F 105 62 −15.32 69
Pure field 2249 1160 ... ...
Poor groups 1086 652 ... ...
Rich groups 1626 937 ... ...
All 6780 3528 ... ...
Note. (1)Structure; (2)total number of galaxies, regardless of their magnitude
(N
gal
);(3)number of galaxies above the survey magnitude limit (N
gal
@
M
rlim);
(4)magnitude limit of each structure; (5)number of galaxies above the
magnitude limit given in (4);(6)number of galaxies in the Kim et al. (2016)
sample.
9
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
work with those listed in the V8k catalog we found that all of
our structures have at least one main counterpart in V8k. This
also applies to the five filamentary structures mentioned above,
that are not in Kim et al. (2016)but are considered in this study.
Indeed, Leo Minor B of this work is mostly matched with the
Crater Cloud in V8k; Serpens with the Serpens Cloud; Draco
with the Bootes cloud; Coma Berenices with the Ursa Major
Southern Spur. The Bootes filament has two main counterparts
in V8k: the Bootes Cloud and the Canes Venatici—
Camelopardalis Cloud.
By matching the V8k galaxy catalog with ours, we also
found 232 galaxies that belong to V8k structures that are not
matched to any of ours, namely Cancer—Leo Cloud, Draco
Cloud, V8k structure ID 322, and Ophiuchus Cloud. These
structures are located along the periphery of the field around
Virgo considered in our study. The presence of these clouds is
not a major concern for our study. They only marginally
contaminate our field sample, as in fact 128 out of the 232 (i.e.,
5.6%)are classified as pure field galaxies in our work (see
Section 4.1).
3.3.2. Radial Density Profiles
In this section we provide an estimate of the width and
density contrast of the filaments, with the goal of better
characterizing these overdense structures. Following Lee et al.
(2021), we investigate how the number density of filament
galaxies depends on the distance to the spine, and we calculate
the density of galaxies in cylindrical shells as a function of 3D
distance from the filament spine. Average densities ρat a
distance rfrom the filament spine are calculated within
cylindrical volumes V=πL[(r+δr)
2
−(r−δr)
2
]=4πLδras
() ()() ()rdd
pd
=<+ - <-
rNrrNrr
Lr4,2
gal gal
where Lis the length of the filament (see Table 2). We increase
the radius from 0.2 to 6 h
−1
Mpc in increments of 0.2 h
−1
Mpc,
while we choose δr=0.1 h
−1
Mpc. We show the resulting
density profiles in Figure 7. The filaments span a range of
densities (the yrange of individual plots varies to improve
readability). When comparing densities within ∼1h
−1
Mpc
from the spine, the Ursa Major Cloud is the densest filament
and the Serpens Filament is the least dense.
Almost all profiles show a decrease in galaxy density as
distance increases. The profiles describing some filaments
flatten out at r>3h
−1
Mpc (e.g., the Leo B, Coma Berenices,
Leo Minor B), while others continue to decline over the full
range of the radii probed (e.g., the Ursa Major Cloud, and
Virgo III). These results suggest that the region around the
Figure 7. Galaxy density versus distance from the filament spine. Dashed lines show the exponential fit, see text for further details.
10
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
filament spine is indeed where the clustering of galaxies is
stronger, and they strengthen the characterization of the
filament skeletons adopted in this work.
The W–M Sheet is an exception: it appears to be the only
structure for which the density is not clearly declining with
distance. Omitting the first two points at small radii that have
large uncertainties due to small number statistics, the profile is
fairly flat up to 1.5 h
−1
Mpc and declines at larger distances.
This finding is not surprising, given the planar distribution of
galaxies in this structure (e.g., Kim et al. 2016).
We fit the density profiles of each filament as a function of
perpendicular distance from the spine, r, with an exponential
law:
() ()r=-+
⎜⎟
⎛
⎝⎞
⎠
ra r
rbexp , 3
0
where ais the best-fit central density at r=r
0
, above the field
value; bis the best fit for the field density at large scales
r?r
0
; and r
0
is the exponential scale width of the filament.
The best-fit parameters are reported in Table 6, while Figure 8
shows the exponential scale width r
0
and the central density
contrast as a function of filament length. The central density
contrast is defined as the density enclosed within r<1h
−1
Mpc
divided by the best-fit value of b. On average we find
r
0
=(0.9 ±0.7)h
−1
Mpc. We report here the median value
along with the rms dispersion around the median.
20
Interest-
ingly, the long filaments tend to have small values of r
0
<1h
−1
Mpc and low-density contrasts <5, whereas shorter filaments
with L<17 h
−1
Mpc have a larger dispersion and reach higher
values for both r
0
and the density contrast.
This analysis is based on the full catalog of ∼7000 sources.
This allows us to better recover the structural parameters of the
filaments with a maximum signal-to-noise ratio. If we repeat
the analysis using the magnitude-limited sample, we obtain
similar results, but strong shot noise in several radial bins
prevents us from deriving robust fits. By using the full catalog
we might be biased toward observing the highest number
densities for the nearest filaments. For example, the Ursa Major
Cloud is nearby and very rich. Similarly, other closer filaments
such as Leo Minor and Canes Venatici show high central
densities. However, our key estimated parameters such as the
density contrast and the scale length r
0
are fairly independent of
the exact galaxy selection, as they are determined relative to the
field density value, which is set at large radii r?r
0
.
Our results are consistent with the theoretical expectations of
Galárraga-Espinosa et al. (2020)for the local universe, who
find that long filaments are thinner and less dense than shorter
ones. Compared to the best fits by Lee et al. (2021)for the
major Virgo III, Canes Venatici, Leo II A, Leo II B, Leo Minor,
and NGC 5353/4filaments, we find smaller central densities
and higher scale length parameters. They found r
0
<1h
−1
Mpc
for all their filaments, which may be due to the fact that they
adopted a different approach. In particular, they used a moving
bin along the radial direction to estimate the density and fit the
profile fixing b=0h
3
Mpc
−3
.
4. Additional Environmental Metrics
4.1. Groups and Field around Virgo
In Section 3.3 we focused on the determination of the
filaments, neglecting the presence of other structures, e.g.,
galaxy groups. It is likely that groups are present both within
filaments and in other field regions. As a consequence, galaxies
outside of the Virgo cluster or the identified filaments are not
necessarily purely field galaxies. To identify galaxy groups
within our sample, we match our catalog to the environmental
catalog from Kourkchi & Tully (2017). They characterized
galaxy groups in our immediate neighborhood (v
r
<3500
km s
−1
). Their group-finding procedure starts with the most
luminous galaxy and iteratively associates galaxies that fall
within its turnaround radius. The algorithm then proceeds to the
next most luminous galaxy that is not already assigned to a
group, and the process repeats. Their galaxy catalog involves a
compilation of sources taken from the Lyon-Meudon Extra-
galactic Database (LEDA
21
), the 2MASS Redshift Survey,
2MRS11.75 (Huchra et al. 2012), and NED. For each galaxy in
their catalog, Kourkchi & Tully (2017)provide the member-
ship to a group and the properties of the group. Of interest for
our scope is the halo mass of the hosting structure, derived
from the Ks-band luminosity by using the M/Lratios given in
their Equation (8). Therefore, Kourkchi & Tullyʼs catalog
allows us (1)identify galaxies that, regardless of their
membership to any filament, belong to a group; (2)obtain a
“clean”pure field sample made up of galaxies not belonging to
any filaments nor associated with groups of two or more
galaxies; and (3)obtain a halo mass estimate of the hosting
structure for each galaxy in the sample.
We cross-match our galaxy catalog and the catalog of group
galaxies of Kourkchi & Tully using a search radius of 10″, and
we find 5651 matches (83% of the sample). For the 1129
galaxies with no match in the Kourkchi catalog, we assign the
group membership of their closest neighbor in 3D space.
Table 6
Best Fits for the Density Profiles of the Filaments, Based on the Whole Sample
Structure abr
0
(h
3
Mpc
−3
)(h
3
Mpc
−3
)(h
−1
Mpc)
Leo Minor F 2.02 ±0.32 0.15 ±0.26 2.67 ±1.24
Canes Venatici F 2.05 ±0.43 0.52 ±0.05 1.40 ±0.31
Bootes F 2.44 ±0.74 0.32 ±0.03 0.91 ±0.21
Ursa Major Cloud 6.54 ±0.99 0.34 ±0.03 0.90 ±0.09
Leo II B F 1.39 ±0.90 0.42 ±0.03 0.76 ±0.38
Virgo Leo II A F 1.48 ±0.21 0.12 ±0.08 2.21 ±0.59
Virgo III F 4.66 ±0.71 0.11 ±0.02 0.87 ±0.08
Leo Minor B F 2.37 ±8.89 0.35 ±0.04 0.26 ±0.51
W–M Sheet 1.89 ±0.49 0.16 ±0.06 1.42 ±0.36
NGC 5353/4 F 1.00 ±0.47 0.13 ±0.02 0.83 ±0.30
Serpens F 0.79 ±1.84 0.13 ±0.01 0.25 ±0.30
Draco F 1.27 ±0.64 0.05 ±0.01 0.60 ±0.17
Coma Berenices F 5.44 ±7.56 0.31 ±0.01 0.18 ±0.10
Note. The density profile is parameterized as
()
()
r
=-+ra bexp r
r0.
20
Note that r
0
is overall smaller than the value of 2.3h
−1
Mpc that we found in
Paper I. Discrepancies might be due to both the different sample selection and
the different local density estimator adopted. Indeed, Paper Iconsidered only
filament galaxies in a mass-complete sample and the fifth-nearest-neighbor
density estimator. This is a good proxy for the local density, but overdense and
underdense substructures within the filaments tend to increase the scatter of n
5
when plotted vs r. On the other hand, the density in Equation (2)is averaged in
cylindrical shells, so that this observed scatter is limited.
21
http://leda.univ-lyon1.fr/
11
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
We then classify as pure field galaxies those that are isolated
based on Kourkchi & Tullyʼs classification and do not belong
to the Virgo cluster or to any filament. A total of 2249 galaxies
in our catalog (1160 above the magnitude completeness limit)
are pure field galaxies. Regardless of their membership in any
filaments, 1086 (652 above the magnitude limit)galaxies
belong to groups with 2 N
mem
<5, with N
mem
being the
number of group members identified in the Kourkchi & Tully
catalog. Hereafter, we refer to the 2 N
mem
<5 groups as poor
groups. We define rich groups as those with N
mem
5, and we
find that 1626 galaxies (937 above the magnitude limit)belong
to a rich group and are not in the Virgo cluster. The median
(mean)number of members in a group is 8 (15).
Figure 9summarizes the different environments consid-
ered, showing the overlap among the different classes. A
significant fraction (33%)of galaxies in our sample are pure
field galaxies, while the remaining ones are associated with
megaparsec-scale overdense structures: the Virgo cluster
(17%), the surrounding filaments (31%), and groups (40%).
Filaments are a very heterogeneous environment: 20% of
their galaxies are in common with Virgo and are thus
classified as members of both the cluster and a filament
(Section 3.3), 12% of them are also located in poor groups,
and 36% of them are also found in rich groups. It is therefore
essential to distinguish among the different global environ-
ments in which sources live if we are to understand the
impact of these environments on the observed properties of
the galaxy.
We then extract from Kourkchi & Tullyʼs(2017)catalog the
halo mass of the hosting system. We note that ∼20% of our
cluster galaxies are not members of Virgo according to
Kourkchi & Tully (2017)but instead are formally associated
with lower-mass halos, with masses uniformly distributed
down to ()~
MMlog 10
halo . This discrepancy is due to
differences in the cluster membership assignments between
Kourkchi & Tully (2017)and this work, in particular in the
outskirts of the Virgo, where the memberships are more
uncertain. To avoid confusion and to be consistent with the
Virgo membership definition used in this paper, we assign them
the halo mass of Virgo ∼10
15
M
e
(Fouqué et al. 2001;
Kourkchi & Tully 2017).
4.2. Local Density
In the previous sections we focused on a global parameter-
ization of the environment. We now focus on a more local
prescription in terms of local density. For each galaxy in the
catalog, we compute the k-nearest-neighbor density (with
k=5)
22
. This is a widely used nonparametric estimate for the
local environment of galaxies that is largely independent of the
dark matter halo mass (see, e.g., Muldrew et al. 2012 for a
review). We consider only neighbors in the catalog whose r-
band absolute magnitude is M
r
−15.7, the completeness limit
of the survey to avoid biasing our estimates toward lower
values at higher distances.
23
Specifically, local densities are computed in 3D (volume
densities)in the (SGX; SGY; SGZ)Cartesian frame and in 2D
(surface densities)by projecting separations onto the (SGX;
SGZ)plane. The 2D density is evaluated by including galaxies
within a ΔSGY =5.6h
−1
Mpc width, which corresponds to the
2σstatistical uncertainty along the line of sight at the distance
of Virgo (see Figure 2). As outlined in Section 3.1 line-of-sight
uncertainties are in fact of the order of ∼0.1 dex and may affect
our 3D analysis.
To investigate possible biases in the density estimates, we
compare the 2D versus 3D local densities in Figure 10.The
two density estimates are consistent with each other once
the 2D estimates are rescaled for the SGY width to convert
them into 3D densities, i.e., by dividing them by 5.6 h
−1
Mpc. The median logarithmic difference ()-nlog 5,3D
()D=
-
+
nlog SGY 0.06
5,2D 0.28
0.29 yields a negligible bias, well
within the reported 1σconfidence interval. Given that results
obtained with the 2D and 3D local density estimates are
quantitatively in agreement, from now on we will be
considering only the 3D densities.
As previously shown in Figure 3, velocity dispersion can be
as high as a few thousand kilometers per second in the dense
central regions of Virgo, where the gravitational potential is the
highest. Model-corrected distances are thus uncertain in the
Figure 8. Scale length r
0
(left)and central density contrast (right), i.e., the ratio of the density enclosed within 1 h
−1
Mpc to the best-fit value bat large radii, as a
function of filament length.
22
Using other estimators, such as the modified 10th nearest-neighbor density
(Cowan & Ivezić2008), will not affect the results.
23
Note that in Paper Iwe have not applied a magnitude cut to compute local
densities. Therefore, the two measurements of local density are not directly
comparable.
12
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
proximity of Virgo and in particular at the caustics. This was
illustrated also in Figure 2: the scatter between model-corrected
and redshift-independent distances indeed increases at the
distance of Virgo. This results in larger uncertainties for the
local densities of Virgo members with respect to those
estimated for galaxies in less dense environments. To account
for a possible bias, we consider the extreme scenario where all
cluster members are located at the same distance. This yields
3D local densities for Virgo cluster galaxies that are on average
∼0.4 dex higher. By collapsing the line-of-sight depth of the
Virgo cluster into one distance, the associated 3D densities
represent an upper limit. We discuss the implications of this
further in the next sections when referring to local densities for
Virgo members.
5. Comparing the Different Parameterizations of the
Environment
We are now in a position to compare the different metrics
adopted to define the environment: the cluster, filament, and
group memberships; local densities; and the halo masses of the
hosting structure. By looking for possible differences between
the different environments, we can gain insights into the
physical mechanisms acting at different scales.
Figure 11 focuses on the global environment: the left panel
shows the halo mass distribution of the different subsamples. A
correlation between halo mass and environment appears clear:
the different environments span different ranges in halo masses,
and the typical halo mass increases from pure field galaxies—
peaking around M
halo
=10
11
M
e
—to poor groups to rich
groups to the cluster. The separation in halo mass between
poor and rich groups is quite evident and occurs at
M
halo
∼10
12.2
M
e
. This rather clear cut justifies our choice to
use the richness of five members as a threshold to separate poor
and rich groups.
Turning the attention to filaments, which are the focus of our
analysis, we observe that they span a wide range in halo mass
and the distribution is rather flat, suggesting that filaments can
also host or, more generally, be linked to structures of different
halo masses. About ∼400 filament galaxies are also formally
associated with the Virgo cluster halo itself. This is because, as
already mentioned, the Ursa Major cloud and the W–M Sheet
extend up to the Virgo cluster region itself.
To further investigate the connection between filaments and
groups, the right panel of Figure 11 shows the position of the
groups identified by Kourkchi & Tully (2017)overplotted with
the position of the filaments, identified by their spines for the
sake of clarity. Some filaments do not contain any rich groups,
while others clearly include groups, with varying incidence
(from few to 50% of the galaxies). In particular, the NGC
3535/4filament is named for the rich group where the filament
seems to terminate, i.e., the filament knot (Kim et al. 2016).
The Virgo III filament is an alignment of several groups (e.g.,
NGC 5248, 5364, 5506, 5566, 5678, 5746, and 5775)and
terminates to the east with the NGC 5846 group.
24
When investigating galaxy properties in filaments, it is
therefore important to consider the presence or absence of
galaxy groups. We note that the spine of the Ursa Major Cloud
seems very short when compared to the distribution of member
groups presented in Figure 5. This is merely a projection effect,
as the closest point of the Ursa Major cloud to Earth is only 2.6
h
−1
Mpc. At this distance, filament member galaxies, defined
as those within 2 h
−1
Mpc from the spine, are spread over 30°
on the plane of the sky and appear to have a large projected
distance from the southern end of the spine.
Next, we correlate the global and local environments by
investigating the local density distribution in galaxies in
different global environments (Figure 12). Cluster, filament,
and pure field galaxies cover different density ranges, with pure
field galaxies lying preferentially at lower densities and cluster
galaxies at the highest ones. Filament galaxies span an
intermediate range of local densities. This agrees with
predictions from simulations (Cautun et al. 2014)and with
what we already showed in Paper I, though for a smaller
sample of filament galaxies. Nonetheless, there is a nonnegli-
gible overlap among the different distributions, indicating that
Figure 9. Euler–Venn diagram summarizing the distribution of galaxies in the
different global environments.
Figure 10. Comparison between the local number density estimates for the
galaxies in the sample. The x-axis shows the 3D volume number densities. The
y-axis displays the 2D surface densities translated into 3D volume densities,
obtained dividing by the slice width ΔSGY =5.6h
−1
Mpc.
24
http://www.atlasoftheuniverse.com/galgrps/viriii.html
13
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
there are low-density regions in the cluster and relatively dense
regions in the field. The median density of the filament galaxies
is considerably influenced by the Ursa Major Cloud and the W–
M Sheet, which host galaxies simultaneously belonging to both
the clusters and the aforementioned structures, at the high-
density tail of the distribution.
The Ursa Major Cloud and the W–M Sheet are not the only
structures sharing galaxies with other systems: as already
mentioned, other filaments share galaxies with groups of
different richness. It could therefore be possible that the large
density range probed by filaments is driven by the presence/
absence of groups. In Figure 13 we therefore compare the
density distribution of filament galaxies (red histograms)to the
density distribution of group galaxies (blue histograms)that are
also in the filaments, subdivided in bins of halo mass. A shift
toward larger densities when increasing the halo mass is clearly
visible, confirming that filament galaxies at the highest
densities are likely also members of a group.
Finally, we inspect the density distribution of the different
filaments, separately, to determine if overall all filaments
behave similarly or if there is a wide filament to filament
variation. To increase the statistics, for each filament we use its
proper completeness limit (see Table 5)and extract from the
field and cluster samples only galaxies above the same limit
and located up to the same distance. Figure 14 highlights that
different filaments are characterized by different density
distributions, taking into account both the median and the
range in density.
To conclude, the main result of this section is that even
though the local and global parameterizations of the environ-
ment agree qualitatively with each other, there is no clear one-
to-one correlation between the two. This demonstrates that
contrasting the variation of galaxy properties as a function of
the global and local environment separately is important in
identifying the acting physical mechanisms.
6. Properties of the Galaxies in Different Environments
In this section we provide an overview of the properties of
galaxies located in different environments. We consider the
de Vaucouleurs morphological parameter (simply called
morphology from now on, Section 6.1)and the presence of
bars (Section 6.2). These parameters are taken from the
HyperLeda catalog. Above the completeness magnitude limit
M
r
=−15.7, 3485/3530 galaxies have a value of morph-
ology, and 3450/3530 have information on the presence or
absence of a bar.
The HyperLeda catalog also provides information on the
position angle of each galaxy. Similarly to Paper I,wemeasure
the projected orientation θ
alignment
between the major axis of each
filament galaxy and the direction of the filament spine, estimated
at the point of minimum distance from the galaxy. The alignment
is thus the galaxy position angle, 0°θ
alignment
90°, with
Figure 11. Left: halo mass distribution for galaxies in the different environments, using the group memberships and halo mass distributions from Kourkchi & Tully
(2017). Right: spatial distribution of groups and filaments around the Virgo galaxy cluster. Shaded gray points represent all groups in the velocity range
500 <v
r
<3300 km s
−1
according to Kourkchi & Tully (2017). Lines represent the filament spines; colored points represent the groups that share galaxies with the
corresponding filament, plotted with the same color. The size of the points scales as the halo mass.
Figure 12. Volume 3D number density distribution for galaxies in the filaments
(red), in the Virgo cluster (green), and in the field (gray)for galaxies above the
absolute magnitude completeness limit. The reported errors are standard
deviations, while the error of the means is much smaller, of the order of
0.05 dex at most. The firebrick cross and error bar show the median density for
filaments when the W–M Sheet and the Ursa Major Cloud are removed from
the filament sample.
14
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
respect to the projected orientation of the filament in the plane of
the sky. In Section 6.3 we search for possible features in the
alignments of galaxies in the filaments.
6.1. Morphologies
We investigate the morphological properties of galaxies as a
function of their global environments (cluster, filaments, groups,
field)and the associated local parameterization in terms of local
densities. We will distinguish galaxies between early type (de
Vaucouleurs morphological type T<0, ET)and late type
(T0, LT).
6.1.1. Morphology and the Environment
Figure 15 shows the incidence of each morphological type in
the different global environments. As seen in the left panel,
there is a clear dichotomy in the morphology of cluster and
pure field galaxies, which have preferentially early- and late-
type morphology, respectively.
Galaxies in poor groups follow quite closely the trend of the
pure field galaxies, while overall rich groups and filaments
have intermediate behaviors, with an excess of ET galaxies
with respect to the pure field, and an excess of LT galaxies with
respect to the cluster.
To understand if the trends in filamentsdependonthe
presence/absence of massive groups within them or if filaments
are truly a site of transformations, in the right panel of Figure 15
we compare the morphological distribution of galaxies only
belonging to filaments to those belonging simultaneously to a
filament and a rich group. Filament galaxies that are not in rich
groups exhibit a bimodal morphological distribution: galaxies
haveeitheraveryETorLTmorphology, while intermediate
values are less favored. The observed excess of ET galaxies with
respect to the pure field suggests that filaments induce a
morphological transformation, even when groups within them
are not included.
In contrast, galaxies of rich groups, either in filaments or not,
show a fairly uniform distribution in morphological type,
suggesting that rich groups act as the main driver for the
suppression of the LT galaxy excess that is typical of the pure
field. The fraction of the earliest type is the highest when galaxies
arebothinrichgroupsandfilaments, suggesting that the
combination of the two environments promotes transformations.
When considering poor groups, we verified that differences
between galaxies in both filaments and groups and only in
filaments disappear, indicating that poor groups do not play a
major role in inducing morphological transformations.
The results above highlight that the dependence of morphology
on the global environment is complex. This is particularly true for
filaments, which span four orders of magnitude in local density.
We therefore look for any morphological trends as a function of
local density. Figure 16 shows the median morphological T-type
plotted against the median local density for each filament,
separately. Cluster and pure field values are shown for
comparison. Overall, even though the scatter is large, the two
quantities are anticorrelated: denser structures tend to be
dominated by ET galaxies. Filaments are intermediate between
the pure field and the cluster, and a large filament-to-filament
variation is detected on both axes. A few structures, i.e., Leo
Minor, Canes Venatici, Leo Minor B, and Serpens, show almost
no ET galaxies. These are filaments with only a few groups
(Figure 11, right)and with the lowest average densities. In
contrast, Virgo III, Ursa Major Cloud, and the W–M Sheet have
on average the highest local densities, higher fractions of ET
galaxies, and are rich in groups. We remind the reader that Virgo
III is an alignment of several groups, while both the W–M Sheet
and the Ursa Major Cloud are connected to the Virgo cluster itself.
This may explain at least partially their higher local densities and
the prevalence of ET galaxies.
To conclude, the above results show that the ET galaxies are
largely present already in filaments, which support the scenario
that morphological transformations may occur well before
galaxies fall into the cluster core.
6.1.2. Morphological Fractions in the Different Environments
We now quantify the variation of the morphological fraction
with the environment, more specifically in terms of the LT
fraction, i.e., the number of galaxies with an LT morphology over
the total. The left panel of Figure 17 shows that, considering the
different global environments, the fraction monotonically
increases from the cluster (∼40%)to the pure field (>80%),
while filaments have an intermediate fraction (60%). Interestingly,
galaxies that are both in groups and filaments have a lower
probability of being LT than both galaxies in groups only and
Figure 13. Volume number density distribution for galaxies in groups of
different halo masses and in filaments (blue histograms). Red histograms show
the overall distribution for filament galaxies. The solid vertical lines show the
median values of the distributions, while the error bars show the 1σ
uncertainties.
15
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
Figure 14. Volume number density (n
5,3D
)distribution for galaxies in each filament separately (red). For each filament, the proper absolute magnitude limit (see
Table 5)has been adopted to increase the statistics. For comparison, the distributions of galaxies in the Virgo cluster (green)and in the field (gray)are also reported,
above the same completeness limit and limiting the sample to the same velocity. Vertical lines represent median values, horizontal lines the standard deviation,
representing the scatter of the distribution.
Figure 15. Fraction of galaxies of different morphological types in the different environments, as described in the legend. A small arbitrary horizontal shift has been
applied to the points for the sake of clarity.
16
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
sources in filaments only. This result again points to the scenario
according to which both filaments and groups affect morphology,
separately, and their effect is amplified for galaxies simultaneously
in both environments. We verified that we obtain similar results
when considering the halo mass of the hosting structure, with the
fraction of LT decreasing with increasing halo mass.
We are now in the position to investigate the so-called
morphology–density relation (Dressler 1980). This relation was
first established for clusters only, then groups (Postman &
Geller 1984), and we now inspect it also in other global
environments (central and right panels in Figure 17)to determine
which environmental definition plays the major role. In each
global environment taken separately, we see a decline of the LT
fraction with increasing density. Nonetheless, the global environ-
ment does play a role in shaping the LT fraction: at a fixed
density, the LT fraction increases from cluster, rich groups,
filaments, poor groups, and to the pure field. As discussed in
Section 4.2, local densities in Virgo are more uncertain than for
the other global environments. We have therefore computed the
cluster morphology–density relation using the density estimates
obtained assuming that all cluster galaxies are at the exact same
distance. This provides a conservative upper limit on the local
density estimates, as the distance along the line of sight between
cluster galaxies is artificially set to zero. This compression of
distances yields local densities that are ∼0.4 dex higher, on
average, than the actual estimate for the local density of cluster
galaxies. The right border of the green area in Figure 17 shows the
relationship derived when using the upper limits on local density.
Even assuming the upper limits as true values for the local density
of cluster galaxies, their associated LT fractions only tentatively
reach those of filament galaxies. This result shows that
uncertainties associated with the local densities of cluster galaxies
do not impact our results: the observed differences between the
global environments considered remain.
In the right panel of Figure 17 we look for other possible
differences when considering filament and rich group galaxies
in all possible combinations. While these environments showed
different morphology distributions (Figure 15 right), these
differences disappear in the LT fraction versus density plot.
This suggests that the overall density–morphology relation is
similar for groups and filaments, even if there are measurable
differences in the morphological composition of their galaxy
populations.
Finally, we investigate the dependence of LT fraction on the
distance to the cluster and to the filament spines (plots not
shown). The LT fraction of filament, field, and group galaxies
is flat, up to the largest clustercentric distances (∼30 h
−1
Mpc).
A similarly flat behavior is observed as a function of the
distance to the filament spines for filament members. In
particular, to appreciate a trend (if any)we should reach larger
distances from the filament spine than 2 h
−1
Mpc, i.e., the
radius up to which filament membership is assigned. This is a
consequence of the fact that at larger distances we have the
strongest density contrast with respect to the central regions
close to the filament spines.
6.2. Bars
We now investigate the presence of bars in our sample. In
this subsection only, we conservatively exclude both elliptical
and irregular galaxies because they typically do not show
evidence of bars. We thus limit ourselves to lenticular and
spiral galaxies (i.e., −3T8), and we consider as barred
galaxies those sources that are classified as barred by
HyperLeda.
In Figure 18 we investigate the bar fraction as a function of
the global environment (left panel)and as a function of the
local density (center)and morphology (right), considering each
filament separately. There is a light but rather systematic
decrease in the bar fraction from the cluster, to filaments, to the
pure field. This trend might be at least partially due to local
density. As illustrated in the central panel, a mild trend toward
a higher bar fraction for increasing local density is observed.
Filaments are intermediate between the cluster and the pure
field, showing bar fractions of ∼0.35 and ∼0.6 for the field and
cluster, respectively. In contrast, as shown in the right panel of
Figure 18, we do not observe any clear trend of the bar fraction
as a function of the average/median morphology.
In Paper Iwe showed that the fraction of galaxies with star
formation below the main sequence monotonically increases in
filaments with increasing local density. The observed trend for
the bar fraction as a function of local density could thus be
related to the fact that the presence of bars may favor the
cessation (quenching)of star formation, as suggested by a
number of studies (e.g., James & Percival 2016; Newnham
et al. 2020; Fraser-McKelvie et al. 2020).
A large scatter is nonetheless observed when comparing the
different filaments, with the nearby ones preferentially showing
the highest bar fractions. An example is the nearby Leo Minor
filament, which has a very high bar fraction of ∼0.8 and low
average density, while its galaxy population is mostly composed
of LT galaxies. Note, however, that these results are based on only
7 galaxies, while the number of barred galaxies in the other
filaments range from 15 to 82. In addition, we note that the bar
identification is a very delicate task and that the bar detection in
HyperLeda has been attemptedonlyforasmallfractionof
galaxies, mostly those larger than 1′of diameter, which may cause
a possible bias in the determination of the bar fraction.
Figure 16. Median morphological parameter for filaments (colored stars), pure
field (gray point), and the Virgo cluster (green square). Error bars represent 1σ
dispersion. The color code for filaments is the same as in Figure 5.
17
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
Furthermore, the classification likely comes from optical images,
while bars are better seen in the infrared (Eskridge et al. 1999).
Overall, from our analysis, we find that bars are found in
56% of lenticular and spiral galaxies in filaments. The fractions
are distributed as follows: 41% (39/95)for lenticulars
(−3T<−1), 59% (73/124)for ET spirals (−1T<3),
and 60% (154/257)for LT spirals (3T8). The total
fraction of galaxies with strong or weak bars should be closer
to 2/3 as SA, SAB, and SB galaxies are in proportions of 1/3
each (e.g., Eskridge et al. 1999). It is thus likely that some of
the galaxies in our sample are misclassified as nonbarred as
a consequence of the observational uncertainties mentioned
above.
Nevertheless, our bar fractions are fairly in agreement with
those found for galaxies in the local universe, in the range
∼(45–60)%(Marinova & Jogee 2007; Reese et al. 2007; Barazza
et al. 2007). In particular, Aguerri et al. (2009)considered the
redshift range 0.01 <z<0.04 and found fractions equal to 29%,
55%, and 54% for lenticulars, and ET and LT spirals, respectively.
To derive these fractions the authors analyzed the r-band images
of a large sample of galaxies in SDSS down to an absolute
magnitude limit of M
r
=–20, a magnitude limit that is brighter
than what we use in this work. By using the same magnitude cut
adopted by the authors we obtain even higher fractions for all
considered classes, with an overall bar fraction of 71%. These
differences highlight the difficultyinassessinganabsolutebar
fraction that is independent of the sample selection, the images
used, and the method adopted to detect the bars.
6.3. Galaxy Alignments with Respect to the Filament Spines
We conclude the overview of galaxy properties by
investigating the alignment of filament galaxies with respect
to the filament spine.
Overall, for each of the filaments, we verified that the
distribution of θ
alignment
is fairly uniform, with mean alignments
around 45°. Previous studies (Tempel et al. 2013; Tempel &
Libeskind 2013b; Hirv et al. 2017; Codis et al. 2018; Chen
et al. 2019; Welker et al. 2020; Kraljic et al. 2021)found
that the spin axis of ET and LT galaxies is preferentially
perpendicular and parallel, respectively, to the filaments. The
expected difference can be ultimately related to the different
Figure 18. Bar fraction as a function of the environment (left), local density (center), and morphology (right). Only lenticular galaxies and normal spirals are
considered; see text for details. For the left panel, the different environments are reported as in Figure 17 (left). In the central and right panels, filaments (colored stars),
pure field (gray point), and the Virgo cluster (green square)are distinguished. The color code for the filaments is the same as in Figure 5.
Figure 17. LT fraction as a function of different environments (left)and local density (center, right). Different environments are considered in the left panel: cluster
galaxies (CL),filament galaxies (F), pure field galaxies (PF), galaxies in rich (RG)and poor (PG)groups, as well as a combination of these classes, at intermediate
environments. In the central panel, the right border of the green dashed area defines the conservative upper limit to the local density for Virgo cluster galaxies. A small
arbitrary horizontal shift has been applied to the points for the sake of clarity.
18
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
assembly histories of ET and LT galaxies. ET galaxies are
thought to be predominantly formed via major mergers. During
these events, the rotation axis of the resulting galaxy tends to be
perpendicular to the merger direction. For LT galaxies, the
assembly primarily occurs via the winding of flows, and the
alignment of angular momentum with the filament spine is
related to the regions outside filaments, namely sheets, where
most of the gas is falling in from (Tempel & Libeskind 2013b).
In Figure 19 we therefore inspect the alignment distributions for
LT and ET galaxies, separately. Data distributions are shown in
terms of violin plots, which give the probability density of the data
at different values, smoothed by a kernel density estimator. Unlike
bar graphs with means and error bars, violin plots show the
distribution of all data points. The shape of the violin displays the
frequencies of values: the thicker part of the violin shape means
that the values in that y-axis section of the violin have a higher
frequency, and the thinner part implies lower frequency. Violin
plots also highlight the maximum extension of the data, and the
presence of different peaks, their position and relative amplitude.
The maximum width of each violin is set the same for all galaxies,
for display purposes.
While for LT galaxies the average alignments scatter around
45°for all filaments, for ET galaxies we do see a higher
filament by filament variation, with median θ
alignment
values
ranging from ∼20°to 80°. This large scatter may be due to the
limited number of ET galaxies in each filament, which is
reported in parentheses above each violin in the Figure.
We did not find any statistically significant difference when
comparing the overall distribution of θ
alignment
of ET and LT
galaxies with the Kolmogorov–Smirnov test. This is partially at
odds with the aforementioned studies; however, the absence of
a significant difference may be due to uncertainties associated
with the determination of the alignment angle, as the analysis is
done in projection and relies on the position angles of the
galaxy and the filament, estimated locally, which are both
uncertain. Similarly, no difference has been found in θ
alignment
,
when considering barred and nonbarred galaxies, separately.
No trend of θ
alignment
as a function of distance to the filament
spine has been found.
7. Summary and Conclusions
We have presented a comprehensive catalog of galaxies
extending up to ∼12 virial radii in projection from the Virgo
cluster, with the intent of characterizing the complex network of
filamentary structures around Virgo and investigating the role of
filaments in galaxy evolution. We select spectroscopically
confirmed galaxies from HyperLeda, the NASA Sloan Atlas,
NED, and ALFALFA, assembling a sample of galaxies in the
region 100°<R.A. <280°,–1.3°<decl. <75°,withrecession
velocities in the range 500 <v
r
<3300 km s
−1
. These cuts
ensure that both Virgo and its main filaments in the northern
hemisphere are included. The final catalog contains 6780
galaxies, 3528 of which are brighter than the absolute magnitude
limit M
r
=−15.7 (+M
3
r, Blanton et al. 2005).
Figure 19. Violin plots of the alignments θ
alignment
for ET (top)and LT (bottom)galaxies in filaments. For each filament, medians and the interquartile ranges are also
shown with circles and thick bars, respectively. We report in parentheses the number of galaxies in each filament.
19
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
To characterize the environment around Virgo, we adopt a
number of parameterizations that trace different scales. By
exploiting a tomographic approach, we recover 13 filaments,
spanning several megaparsecs in length.
We then assign filament memberships, relying on the 3D
distance of the galaxies from the filament spines, which we release
for all 13 considered filamentary structures. We also identify the
cluster members both in the 3D SG coordinate frame and also
consider the cluster region in phase space.
To further characterize the environments of our catalog
galaxies, we match our sample to Kourkchi & Tullyʼs(2017)
group catalog to select galaxies in groups and extract for each
galaxy of the sample the halo mass estimate of the hosting
structure. Finally, we quantify the local environment using
surface (2D)and volume (3D)local densities in terms of the
fifth-nearest neighbors. We make available the catalogs of
galaxies and of the aforementioned environments.
We then characterize galaxy morphology and spin alignment
of galaxies in filaments and discuss the different parameteriza-
tions of the environment. The main results of our analysis are:
1. By fitting an exponential model to the distribution of
galaxies, averaged in cylindrical shells around each filament
spine, we find that long >17 h
–1
Mpc filaments have low
characteristic radii r
0
<1h
−1
Mpc (along the direction
perpendicularly to the filament spine)and the lowest density
contrasts with respect to the field. Shorter filaments have a
larger range of values of both the density contrast and
characteristic radius and extend to higher values in each.
2. Filament galaxies span a wide range of ∼4 dex in both
local density and halo mass of the hosting structure (e.g.,
group). Values range at the low end from those typical of
the field to values found in the Virgo cluster at the high
end. The high dispersion found for the filaments is
ultimately due to the large filament to filament variation
and to the fact that some filaments are very rich in groups,
while others are poorer.
3. A decline of the LT fraction with increasing local density
is observed in all considered global environments (field,
filaments, groups, and cluster).Atfixed local density,
filaments appear to be an intermediate environment
between the field and the cluster, with a decline
resembling that of rich groups. The local density alone
is thus not sufficient to explain the dependence of the LT
fraction with the megaparsec-scale environment.
4. The average fraction of barred galaxies decreases from
the highest-density regions of the cluster to the field at the
lowest density. Filaments show an intermediate and broad
range in the fraction of barred galaxies, with a large
filament to filament variation, which reflects the large
dispersion for filament galaxies observed also in local
density and morphology.
5. We find no clear dependence of the projected orientation
of the galaxy major axis with the filament spine for either
ET or LT galaxies. Similarly, we did not find any clear
trend for the considered properties of filament galaxies as
a function of their distance to the spines. However, it is
important to note that we only consider filament members
to be those galaxies closer than 2 h
−1
Mpc from the
filament spine. While this radius allows us to minimize
contamination from field galaxies, it does make it hard to
assess whether trends would exist if we included galaxies
at larger distances.
The authors thank the hospitality of the International Space
Science Institute (ISSI)in Bern (Switzerland)and of the
Lorentz Center in Leiden (Netherlands). Regular group meet-
ings in these institutes allowed the authors to make substantial
progress on the project and finalize the present work.
G.C. acknowledges financial support from the Swiss
National Science Foundation (SNSF). B.V. acknowledges
financial contribution from the grant PRIN MIUR 2017
n.20173ML3WW_001 (PI Cimatti)and from the INAF main-
stream funding program (PI Vulcani). R.A.F. gratefully
acknowledges support from NSF grants AST-0847430 and
AST-1716657. G.H.R. acknowledges support from NSF-AST
1716690.
This research has made use of the NASA/IPAC Extra-
galactic Database (NED)which is operated by the Jet
Propulsion Laboratory, California Institute of Technology,
under contract with the National Aeronautics and Space
Administration. We acknowledge the usage of the HyperLeda
database.
25
This research made use of Astropy,
26
a community
developed core Python package for Astronomy (Astropy
Collaboration et al. 2013,2018), matplotlib (Hunter 2007),
and TOPCAT (Taylor 2005).
Appendix
Catalogs
With this paper, we release a number of catalogs: the main
galaxy catalog, the catalog of the environmental properties, and
the catalog with the filament spines. The main galaxy catalog is
shown in Table 7for a subsample of 10 galaxies. The table is
presented in its entirety in the online version of the article. The
columns indicate:
1. Column (1)—VFID, a unique serial number, with
galaxies sorted by decl. from north to south;
2. Columns (2)and (3)—R.A. and decl. at epoch J2000 (in
degrees);
3. Column (4)—v
r
heliocentric velocity (units of km s
−1
);
4. Column (5)—V
cosmic
cosmic recession velocity (units
of km s
−1
)obtained from a redshift-independent distance
from Steer et al. (2017)when available or from V
model
as
described in Section 3.1;
5. Column (6)—V
model
model recession velocity (units
of km s
−1
)obtained from the Mould et al. (2000)model,
as described in Section 3.1;
6. Column (7)—HyperLeda name;
7. Column (8)—NED name;
8. Column (9)—PGC ID;
9. Column (10)—NSAID from the v0 catalog;
10. Column (11)—NSAID from the v1 catalog;
11. Column (12)—Arecibo Galaxy Catalog (AGC)name;
12. Column (13)—Boolean flag, where True indicates that
the galaxy has a CO observation from Paper I;
13. Column (14)—Boolean flag, where True indicates that
the galaxy is in the ALFALFA α.100 catalog (Haynes
et al. 2018).
Galaxy environmental properties are listed in Table 8for a
subsample of 10 galaxies, while the table for the total sample is
given in the online version of the article. The columns indicate:
25
http://leda.univ-lyon1.fr
26
http://www.astropy.org
20
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
Table 7
Main Catalog with Cross IDs
VFID R.A. Decl. v
r
v
cosmic
v
model
HL Name NED Name PGC NSA V0 NSA V1 AGC CO A100
(deg, J2000)(deg, J2000)(km s
−1
)(km s
−1
)(km s
−1
)
(1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)
3000 165.930807 28.88713 708 1189 644 NGC 3510 NGC 3510 33408 100677 472983 6126 False True
3001 149.190405 28.82596 510 687 495 UGC 05340 UGC 05340 28714 136251 623560 5340 False True
3002 194.776643 28.81178 998 1177 1177 PGC 1846725 WISEA J125906.48 +284842.6 1846725 89104 427578 −999 False False
3003 157.778262 28.79659 1425 1732 1732 NGC 3265 NGC 3265 31029 107764 497691 5705 True True
3004 129.582068 28.78993 2669 2842 2842 PGC 3095094 WISEA J083819.66 +284723.7 3095094 135383 622813 −999 False False
3005 181.303258 28.78191 3153 3395 3395 UGC 07072 UGC 07072 38268 102495 478264 7072 False True
3006 250.089288 28.76552 976 1291 1291 SDSS J164021.43 +284555.9 SDSS J164021.43 +284555.9 4123676 69715 343115 262737 False True
3007 225.279839 28.76086 1821 2158 2158 SDSS J150107.16 +284539.2 WISEA J150107.08 +284539.7 4443809 −999 −999 733373 False True
3008 142.246120 28.75796 1228 1459 1459 PGC 1845056 SDSS J092859.06 +284528.5 1845056 84921 410169 194058 False True
3009 128.807646 28.75335 2052 3525 2244 UGC 04482 UGC 04482 24104 156791 647654 4482 False True
Note. Galaxies without a corresponding ID in columns 9–12 are denoted as −999.
(This table is available in its entirety in machine-readable form.)
21
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
Table 8
Environmental Properties of Catalog Galaxies
VFID SGX SGY SGZ n
5,2D
err(n
5,2D
)n
5,3D
err(n
5,3D
)Nearest Filament D
Filament
2D D
Filament
3D
Filament
Memb. Group Cluster
Pure
Field
(h
−1
Mpc)
(h
−1
Mpc)
(h
−1
Mpc)(h
2
Mpc
−2
)(h
2
Mpc
−2
)(h
3
Mpc
−3
)(h
3
Mpc
−3
)(h
−1
Mpc)(h
−1
Mpc)
(1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)
VFID3000 2.0 11.3 −3.1 1.9 0.9 0.2 0.1 Coma_Berenices 1.2 2.0 0 2 0 0
VFID3001 2.0 5.8 −3.2 0.4 0.2 0.1 0.0 Leo_Minor 1.4 1.5 1 0 0 0
VFID3002 0.3 11.6 1.7 0.9 0.4 0.2 0.1 Canes_Venatici 1.4 1.4 1 0 0 0
VFID3003 3.9 15.7 −6.3 4.5 2.0 0.5 0.2 LeoII_B 0.1 1.9 1 2 0 0
VFID3004 12.4 18.3 −17.9 0.4 0.2 0.1 0.0 LeoII_A 4.8 7.4 0 0 0 1
VFID3005 2.7 33.8 −1.8 2.7 1.2 0.7 0.3 Coma_Berenices 4.1 4.1 0 0 0 1
VFID3006 0.3 7.1 10.7 0.1 0.0 0.0 0.0 Serpens 5.5 6.8 0 1 0 0
VFID3007 −0.6 17.7 12.3 1.1 0.5 0.1 0.0 Serpens 5.2 6.4 0 0 0 1
VFID3008 4.9 11.3 −7.7 0.4 0.2 0.0 0.0 LeoII_B 0.4 2.5 0 0 0 1
VFID3009 15.6 22.4 −22.4 0.1 0.0 0.0 0.0 LeoII_A 10.3 14.2 0 1 0 0
(This table is available in its entirety in machine-readable form.)
22
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
1. Column (1)—VFID, galaxy unique serial number;
2. Columns (2)–(4)—SG X, Y, and Z coordinates, com-
puted as described in Section 3.1;
3. Columns (5)and (6)—local surface number density and
1σPoisson uncertainty computed as described in
Section 4.2;
4. Columns (7)and (8)—local volume number density and
1σPoisson uncertainty computed as described in
Section 4.2;
5. Column (9)—Name of the nearest filament;
6. Column (10)—2D distance of the galaxy from the nearest
filament;
7. Column (11)—3D distance of the galaxy from the nearest
filament;
8. Column (12)—filament member flag, where 1 indicates
that the galaxy is a filament member, i.e., within 2h
−1
Mpc from the nearest filament spine.
9. Column (13)—group membership flag, according to the
group definition by Kourkchi & Tully (2017): 0 means
the galaxy is not a member of a group, 1 means the
galaxy is a member of a poor group (2N<5), and 2
means the galaxy is a member of a rich group (N5), see
text for details;
10. Column (14)—cluster membership flag as described in
Section 3.2, where 1 indicates that the galaxy is a cluster
member;
11. Column (15)—pure field galaxy flag, obtained as
described in Section 4, where 1 indicates that the galaxy
is a pure field galaxy.
ORCID iDs
Gianluca Castignani https://orcid.org/0000-0001-6831-0687
Benedetta Vulcani https://orcid.org/0000-0003-0980-1499
Rose A. Finn https://orcid.org/0000-0001-8518-4862
Francoise Combes https://orcid.org/0000-0003-2658-7893
Pascale Jablonka https://orcid.org/0000-0002-9655-1063
Gregory Rudnick https://orcid.org/0000-0001-5851-1856
Dennis Zaritsky https://orcid.org/0000-0002-5177-727X
Kelly Whalen https://orcid.org/0000-0002-8571-9801
Gabriella De Lucia https://orcid.org/0000-0002-6220-9104
Vandana Desai https://orcid.org/0000-0002-1340-0543
John Moustakas https://orcid.org/0000-0002-2733-4559
References
Abazajian, K. 2008, ApJS,182, 543
Aguerri, J. A., Méndez-Abreu, J., & Corsini, E. M. 2009, A&A,495, 491
Alpaslan, M., Robotham, A. S., Obreschkow, D., et al. 2014, MNRAS,
440, L106
Aragon-Calvo, M. A., Platen, E., van de Weygaert, R., & Szalay, A. S. 2008,
ApJ,723, 364
Astropy Collaboration, Robitaille, T. P., Tollerud, E. J., et al. 2013, A&A,
558, A33
Astropy Collaboration, Price-Whelan, A. M., Sipőcz, B. M., et al. 2018, AJ,
156, 123
Barazza, F. D., Jogee, S., & Marinova, I. 2007, ApJ,675, 1194
Beygu, B., Peletier, R. F., van der Hulst, J. M., et al. 2017, MNRAS,464, 666
Biviano, A., Fadda, D., Durret, F., Edwards, L. O. V., & Marleau, F. 2011,
A&A,532, A77
Blanton, M. R., Kazin, E., Muna, D., Weaver, B. A., & Price-Whelan, A. 2011,
AJ,142, 31
Blanton, M. R., Lupton, R. H., Schlegel, D. J., et al. 2005, ApJ,631, 208
Blue Bird, J., Davis, J., Luber, N., et al. 2020, MNRAS,492, 153
Bond, J. R., Kofman, L., & Pogosyan, D. 1995, Natur,380, 603
Boselli, A., Cortese, L., & Boquien, M. 2014a, A&A,564, 65
Boselli, A., Cortese, L., Boquien, M., et al. 2014b, A&A,564, 67
Boselli, A., Cortese, L., Boquien, M., et al. 2014c, A&A,564, 66
Boselli, A., Voyer, E., Boissier, S., et al. 2014d, A&A,570, 69
Brouwer, M. M., Cacciato, M., Dvornik, A., et al. 2016, MNRAS,462, 4451
Castignani, G., Combes, F., Jablonka, P., et al. 2022, A&A,657, A9
Cautun, M., van de Weygaert, R., Jones, B. J. T., & Frenk, C. S. 2014,
MNRAS,441, 2923
Chen, Y. C., Ho, S., Brinkmann, J., et al. 2016, MNRAS,461, 3896
Chen, Y. C., Ho, S., Tenneti, A., et al. 2015, MNRAS,454, 3341
Chen, Y.-M., Shi, Y., Wild, V., et al. 2019, MNRAS,489, 5709
Chung, A., Van Gorkom, J. H., Kenney, J. D., Crowl, H., & Vollmer, B. 2009,
AJ,138, 1741
Codis, S., Jindal, A., Chisari, N. E., et al. 2018, MNRAS,481, 4753
Courtois, H. M., Pomarède, D., Tully, R. B., Hoffman, Y., & Courtois, D.
2013, AJ,146, 69
Cowan, N. B., & Ivezić, Z. 2008, ApJ,674, L13
Darvish, B., Mobasher, B., Sobral, D., et al. 2015, ApJ,814, 84
Darvish, B., Sobral, D., Mobasher, B., et al. 2014, ApJ,796, 51
Davies, J. I., Wilson, C. D., Auld, R., et al. 2010, MNRAS,409, 102
Dey, A., Schlegel, D. J., Lang, D., et al. 2019, AJ,157, 168
Dressler, A. 1980, ApJ,236, 351
Durbala, A., Finn, R. A., Odekon, M. C., et al. 2020, AJ,160, 271
Eardley, E., Peacock, J. A., McNaught-Roberts, T., et al. 2015, MNRAS,
448, 3665
Eskridge, P. B., Frogel, J. A., Pogge, R. W., et al. 1999, AJ,119, 536
Fadda, D., Biviano, A., Marleau, F. R., Storrie-Lombardi, L. J., & Durret, F.
2008, ApJL,672, L9
Ferrarese, L., Côté, P., Cuillandre, J. C., et al. 2012, ApJS,200, 4
Fouqué, P., Solanes, J. M., Sanchis, T., & Balkowski, C. 2001, A&A,375, 770
Fraser-McKelvie, A., Merrifield, M., Aragón-Salamanca, A., et al. 2020,
MNRAS,499, 1116
Galárraga-Espinosa, D., Aghanim, N., Langer, M., Gouin, C., & Malavasi, N.
2020, A&A, 641, 173
Geach, J. E., Ellis, R. S., Smail, I., Rawle, T. D., & Moran, S. M. 2011,
MNRAS,413, 177
Giovanelli, R., Haynes, M. P., Kent, B. R., et al. 2005, AJ,130, 2598
Goh, T., Primack, J., Lee, C. T., et al. 2018, MNRAS,483, 2101
Guo, H., Zheng, Z., Zehavi, I., et al. 2014, MNRAS,441, 2398
Haynes, M. P., Giovanelli, R., Kent, B. R., et al. 2018, ApJ,861, 49
Hirv, A., Pelt, J., Saar, E., et al. 2017, A&A,599, A31
Hoyle, F., Rojas, R. R., Vogeley, M. S., & Brinkmann, J. 2005, ApJ,620, 618
Huchra, J. P., MacRi, L. M., Masters, K. L., et al. 2012, ApJS,199, 26
Hunter, J. D. 2007, CSE,9, 90
Jaffé, Y. L., Smith, R., Candlish, G. N., et al. 2015, MNRAS,448, 1715
James, P. A., & Percival, S. M. 2016, MNRAS,457, 917
Jasche, J., Kitaura, F. S., Li, C., & Enßlin, T. A. 2010, MNRAS,409, 355
Joeveer, M., Einasto, J., & Tago, E. 1978, MNRAS,185, 357
Karachentsev, I. D., & Nasonova, O. G. 2010, MNRAS,405, 1075
Kim, S., Rey, S.-C., Bureau, M., et al. 2016, ApJ,833, 207
Kim, S., Rey, S.-C., Jerjen, H., et al. 2014, ApJS,215, 22
Kleiner, D., Pimbblet, K. A., Heath Jones, D., Koribalski, B. S., & Serra, P.
2017, MNRAS,466, 4692
Ko, Y., Hwang, H. S., Lee, M. G., et al. 2017, ApJ,835, 212
Kourkchi, E., & Tully, R. B. 2017, ApJ,843, 16
Koyama, Y., Kodama, T., Nakata, F., Shimasaku, K., & Okamura, S. 2011,
ApJ,734, 66
Koyama, Y., Kodama, T., Tadaki, K.-I., et al. 2014, ApJ,789, 18
Kraljic, K., Arnouts, S., Pichon, C., et al. 2018, MNRAS,474, 547
Kraljic, K., Duckworth, C., Tojeiro, R., et al. 2021, MNRAS,504, 4626
Kraljic, K., Pichon, C., Dubois, Y., et al. 2019, MNRAS,483, 3227
Kreckel, K., Platen, E., Aragón-Calvo, M. A., et al. 2011, AJ,141, 4
Kuchner, U., Aragón-Salamanca, A., Pearce, F. R., et al. 2020, MNRAS,
494, 5473
Kuchner, U., Aragón-Salamanca, A., Rost, A., et al. 2021, MNRAS,503, 2065
Kuutma, T., Tamm, A., & Tempel, E. 2017, A&A,600, L6
Laigle, C., Pichon, C., Arnouts, S., et al. 2018, MNRAS,474, 5437
Lang, D., Hogg, D. W., & Mykytyn, D. 2016, The Tractor: Probabilistic
astronomical source detection and measurement, Astrophysics Source Code
Library, ascl:1604.008
Lee, Y., Kim, S., Rey, S.-C., & Chung, J. 2021, ApJ,906, 68
Leroy, A. K., Sandstrom, K. M., Lang, D., et al. 2019, ApJS,244, 24
Libeskind, N. I., Carlesi, E., Grand, R. J., et al. 2020, MNRAS,498, 2968
Libeskind, N. I., van de Weygaert, R., Cautun, M., et al. 2018, MNRAS,
473, 1195
23
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.
Mahajan, S., Raychaudhury, S., & Pimbblet, K. A. 2012, MNRAS,427, 1252
Mahajan, S., Singh, A., & Shobhana, D. 2018, MNRAS,478, 4336
Makarov, D., Prugniel, P., Terekhova, N., Courtois, H., & Vauglin, I. 2014,
A&A,570, 13
Malavasi, N., Aghanim, N., Douspis, M., Tanimura, H., & Bonjean, V. 2020a,
A&A,642, A19
Malavasi, N., Aghanim, N., Tanimura, H., Bonjean, V., & Douspis, M. 2020b,
A&A,634, A30
Malavasi, N., Arnouts, S., Vibert, D., et al. 2017, MNRAS,465, 3817
Marinova, I., & Jogee, S. 2007, ApJ,659, 1176
McLaughlin, D. E. 1999, ApJ,512, L9
Mei, S., Blakeslee, J., Cote, P., et al. 2007, ApJ,655, 144
Mould, J. R., Huchra, J. P., Freedman, W. L., et al. 2000, ApJ,529, 786
Muldrew, S. I., Croton, D. J., Skibba, R. A., et al. 2012, MNRAS,419, 2670
Newnham, L., Hess, K. M., Masters, K. L., et al. 2020, MNRAS,492, 4697
Odekon, M. C., Hallenbeck, G., Haynes, M. P., et al. 2018, ApJ,852, 142
Pintos-Castro, I., Sánchez-Portal, M., Cepa, J., et al. 2013, A&A,558, A100
Porter, S. C., & Raychaudhury, S. 2006, MNRAS,375, 1409
Porter, S. C., Raychaudhury, S., Pimbblet, K. A., & Drinkwater, M. J. 2008,
MNRAS,388, 1152
Postman, M., & Geller, M. J. 1984, ApJ,281, 95
Reese, A. S., Williams, T. B., Sellwood, J. A., Barnes, E. I., & Powell, B. A.
2007, AJ,133, 2846
Riess, A. G., Casertano, S., Yuan, W., Macri, L. M., & Scolnic, D. 2019, ApJ,
876, 85
Rojas, R. R., Vogeley, M. S., Hoyle, F., & Brinkmann, J. 2004, ApJ,617, 50
Rost, A., Kuchner, U., Welker, C., et al. 2021, MNRAS,502, 714
Santos, J. S., Altieri, B., Tanaka, M., et al. 2014, MNRAS,438, 2565
Sobral, D., Best, P. N., Smail, I., et al. 2011, MNRAS,411, 675
Steer, I., Madore, B. F., Mazzarella, J. M., et al. 2017, AJ,153, 37
Taylor, M. B. 2005, in ASP Conf. Ser., 347, Astronomical Data Analysis
Software and Systems XIV, ed. P. Shopbell, M. Britton, & R. Ebert (San
Francisco, CA: ASP),29
Tempel, E., & Libeskind, N. I. 2013a, ApJL,775, L42
Tempel, E., & Libeskind, N. I. 2013b, ApJL,775, 42
Tempel, E., Stoica, R. S., Mart⧹’⧹inez, V. J., et al. 2014, MNRAS,438, 3465
Tempel, E., Stoica, R. S., & Saar, E. 2013, MNRAS,428, 1827
Tifft, W. G., & Gregory, S. A. 1976, ApJ,205, 696
Tully, R. B. 1982, ApJ,257, 389
Tully, R. B., Courtois, H., Hoffman, Y., & Pomarède, D. 2014, Natur,513, 71
Tully, R. B., Courtois, H. M., Dolphin, A. E., et al. 2013, AJ,146, 86
Tully, R. B., Courtois, H. M., & Sorce, J. G. 2016, AJ,152, 50
Tully, R. B., Rizzi, L., Shaya, E. J., et al. 2009, AJ,138, 323
Tully, R. B., Shaya, E. J., Karachentsev, I. D., et al. 2008, ApJ,
676, 184
Vulcani, B., Poggianti, B. M., Moretti, A., et al. 2019, MNRAS,487, 2278
Welker, C., Bland-Hawthorn, J., van De Sande, J., et al. 2020, MNRAS,
491, 2864
Yan, H., Fan, Z., & White, S. D. 2013, MNRAS,430, 3432
Zhang, Y., Yang, X., Wang, H., et al. 2013, ApJ,779, 160
24
The Astrophysical Journal Supplement Series, 259:43 (24pp), 2022 April Castignani et al.