SH #3: Understanding Galaxies         Group Project
Team B:       Large scale structure
in the local Universe
In this activity, you will investigate the distribution of galaxies in the local universe and the
impact of clustering on Hubble's Law locally. Because the distributions of galaxies and dark matter are
not uniform in the nearby universe (i.e. both tend to cluster), the expansion of the universe
locally is not perfectly smooth. Galaxies are accelerated towards overdensities in the mass, especially
clusters. Your task is to understand the importance of "peculiar velocities", the deviations from smooth
Hubble expansion in the local universe, how we can measure measure them, and how they can be used to indicate
of infall onto large scale overdensities in the local universe. Keep in mind that peculiar velocity studies
aren't easy: the universe is a messy place!
The task of this team is to understand the methodology of deriving peculiar velocities and
the amplitude of
Note: Team C is looking into aspects of the Virgo cluster including the distribution (on the
sky and in redshift space) of the galaxies believed to be members of it.
Hint: you can view a bigger version of the images by clicking on them.
Deviations from Hubble expansion
We know that locally, the expansion of the universe is "perturbed" by the fact that galaxies
tend to cluster. As depicted in the cartoon to the right, taken from a publication in the 1980's,
it has been known for some time that the Local Group is falling towards the Virgo Cluster,
even as the local universe expands.
Use this cartoon to understand why the "peculiar velocity" of galaxies nearby
limits our ability to use Hubble's Law to derive distances from velocities.
If we measure the redshift (radial velocity) of a galaxy in the direction of
the Virgo cluster but closer to us than the cluster, would we expect its observed velocity
to be higher or lower than its distance would suggest if we invoked Hubble's Law?
What if the galaxy were located in the direction of Virgo but at a larger
Aaronson et al 1986 ApJ 302, 536
Hubble expansion around the Virgo cluster
To demonstrate the concept, the next diagram shows the results of
a model of the local flow field
in the plane of the Local Supercluster very close to us. It is taken from
a review article by
Davis and Peebles in 1983. Be aware that the details are a bit out of date;
note that the distance of the Virgo cluster they assumed in making
this model was 15 Mpc, and we know it is larger than that (more like 16.7 Mpc).
However, the diagram should be useful to help you understand the local flow field.
The x axis shows the distance from us. The angular scale (in degrees) measures
the angular distance from the direction of the Virgo cluster as viewed from the Sun.
The heavy contours trace the heliocentric velocity expected for a galaxy at a given distance
away from us in different directions. The short dashed circles show the velocities we would expect
if the velocity field were not perturbed, that is, if the Virgo cluster were not there.
Be sure you understand how to interpret this diagram.
Davis and Peebles, 1983, ARA&A 21, 109
Predicting distances using a flow model: a case study
NED gives a long list of
values of "redshift independent distances"; individual distances have to be used with a grain of salt,
because they are a very inhomogeneous and complex set of data which adopt different distances to Virgo,
different values of the Hubble constant, etc. The NED folks try to correct for those things to adopt
a "mean metric distance". What value do they get? How does it compare with the value you predicted
from the model?
The galaxy NGC 4565 is a spiral galaxy viewed almost perfectly edge-on.
It is located at an angular distance of about 13 degrees from
the center of the Virgo cluster.
Use the tools you know about (e.g. NED and SKYDEG) to learn more about this galaxy.
What is the observed heliocentric recessional velocity of NGC 4565?
According to the model velocity field of the Local Supercluster, at what distance would a galaxy seen
14 degrees from the Virgo cluster have a heliocentric velocity equal to that of NGC 4565? Remembering that
the model underestimates the distance of Virgo, use the diagram to predict the galaxy's distance based
on the model and its observed recessional velocity.
In the mid 1950's, Gerard deVaucouleurs noticed (as we shall soon) that
the nearby galaxies within about 40 million galaxies are not distributed
evenly around the sky, but rather the majority of them trace a band around
the sky. The distribution of "nebulae" in a continuous band was actually
noticed by William
Herschel more than 200 years earlier. DeVaucouleurs realized that the
distribution could be explained if the galaxies were mainly found in a flattened
disk-like structure which he referred to as the "supergalactic plane".
Find the NED
the Virgo cluster to find its coordinates in various systems. Write them in decimal degrees
in epoch J2000. Then think about where Virgo must lie in the SGX, SGY, SGZ system.
Galaxies with supergalactic latitude SGB = 0 degrees lie in this plane. Supergalactic
longitude SGL is measured along the plane with its origin (SGL = 0 degrees,
SGB = 0 degrees) lying near Galactic longitude l = 138 degrees and latitude b = 0 degrees
(i.e. in the Galactic plane too).
An alternative system, called the cartesian supergalactic coordinates SGX, SGY, SGZ.
It requires knowledge of distance because its units are in Megaparsecs.
SGX is directed towards the point SGL = 0 degrees, SGB = 0 degrees
and the SGZ axis towards the north supergalactic pole, SGB = +90 degrees. With this
definition, SGX and SGY are in the supergalactic plane, where SGZ = 0 degrees.
Mathematically, we can use a galaxy's equatorial position (R.A. and Dec.) to calculate its
"supergalactic coordinates", supergalactic longitude (SGL) and latitude (SGB). And,
if we know the distance to a galaxy, the
"cartesian supergalactic coordinates" SGX, SGY, SGZ can be assigned using
some fairly simple mathematical formulae:
| R.A. & Dec.||   |
| l, b ||   |
| SGL, SGB ||   |
| SGX, SGY, SGZ ||   |
SGX = distance*cos(SGL)*cos(SGB)
SGY = distance*sin(SGL)*cos(SGB)
SGZ = distance*sin(SGB)
Calculating distances for real: What the Cornell EGG does!
Since one of the main aims of ALFALFA is to take a census of galaxies in the local
universe, members of the Cornell ExtraGalactic Group are always trying to figure
out how best to determine distances to nearby galaxies. It's one of those never-ending
projects that has to be revisted every few years as new and better observational
data, obtained by many people using many different methods, become available.
(We could spend all week talking just about this, but we won't!)
Above, to the right is Karen's flow model velocity field for the supergalactic plane
(SGZ = 0 degrees). The axes are simply labeled X and Y = SGX and SGY.
It is centered on "us" (the Local Group). The contours trace the velocity a galaxy
would have at a certain distance from us. It's more complicated than the diagram
from Davis and Peebles that we used before because it extends further away from
us and includes more "attractors" than just the Virgo cluster.
Find the Virgo cluster in this diagram.
What other features do you see in the flow model? How can you explain them?
We use all of this information and our flow model to explore and understand the distribution of galaxies
in the local universe.
To calculate distance, we use a "hybrid flow model" developed by Karen Masters
(CU PhD 2004, now at the
University of Portsmouth, UK) to calculate the distance to each object. Since we know
that Hubble's Law does not apply locally, Karen's approach to
assigning distances is a bit complicated. It uses "primary distances" where they are
available and has a built-in assignment of some galaxies to their parent groups or
clusters (most notably, the Virgo cluster). In the absence of some redshift-
independent distance measurement, it uses a model of the local velocity field
(predicting the distance to a galaxy of a given observed heliocentric velocity in
a specific direction) that Karen devised by examining all of the galaxies with
known distances. It's not perfect, but it is better than using Hubble's Law.
Notice that Karen's diagram extends to larger distances that the one
Davis and Peebles which we used above.
from Masters, 2004, CU Ph.D. thesis.
Infall of galaxies into the Virgo Cluster
Numerous groups are engaged in the process of using HST observations to measure the distances
of galaxies in redshift-independent ways. The most accurate distances come from measuring
"standard candles" like the tip of the red giant branch (TRGB). An interesting paper for your team
to investigate is
Infall of Nearby Galaxies into the Virgo Cluster as Traced with Hubble Space Telescope
(Karachentsev et al 2014, Ap. J. 782, 4. Here are Figures 1 and 6 from that paper.
Be sure that you understand what these two figures show. How do they indicate evidence
for infall onto the Virgo cluster? How are they different?
Why does Figure 6 show more points at lower distances and only one point at
large ones (in the Virgo background)?
How do the errors in distances obtained using the TRGB method compare to those
obtained with the baryonic Tully-Fisher relation?
Using the Tully-Fisher relation
Peculiar velocity studies require a lot of hard work both in obtaining the observational
dataset, in converting the observational quantities into useful galaxy properties,
in understanding the uncertainties, biases and limitations, and finally, in using
the data to map the deviations from Hubble flow.
Some years ago, some of us conducted a survey of peculiar velocities in the local
universe. You can find it at
SFI++. II. A New I-Band Tully-Fisher Catalog, Derivation of Peculiar Velocities, and Data Set Properties
Springob et al 2007, ApJSupp 172, 599. Note that this predates the SDSS; in fact, we acquired both
the HI dataset and the images needed to measure magnitudes, sizes, inclinations etc. Here is a file containing the final (useful) data for
galaxies in the SFI++ compliation
with recessional velocities less than 12000 km/s.
What is the "I-band"?
What parts of the sky have been sampled by SFI++? What parts have not?
What is the difference btween the Tully-Fisher relation and the baryonic Tully Fisher relation?
Construct the Tully-Fisher diagram for the galaxies. Be sure that the luminosity increases towards
The recessional velocity here is given as "vcmb". What is the "CMB rest frame"?
Plot the relation between distance (horizontal axis) and the recessional velocity (vertical axis)
in the CMB frame. How does the scatter around the mean relation vary with distance? Why?
Infall into the PPS using the Tully-Fisher relation
Simply put, there were not enough galaxies in the region of the PPS for this older survey to detect
infall into PPS. There was a tantalizing
hint, but we never
felt fully confident in the results.
Here is a file containing the
galaxies in the α.70 catalog in the region of APPSS. How many of them are also contained
in the SFI++ sample? Hint: Think carefully: what is the easiest way to match the two catalogs?.
Make the plot of distance versus recessional velocity for the galaxies in
PPS, α.70 and SFI++. If there were galaxies infalling into PPS with a peak velocity of
1000 km/s (and the data were otherwise perfect), what would this diagram look like?
Use what you have learned to explain to the other UATers what the APPS survey will (hopefully) do
.... in the future!
Other things to consider
Look up some more information about the clusters, and perhaps
find some interesting images of them and their notable galaxies.
What other issues/questions are raised in your mind?
Last modified: Sat Jun 11 10:03:46 EDT 2016 by martha