The First-Ever Detection of the Kinematic Sunyaev-Zel’dovich Effect

Title: Detection of Galaxy Cluster Motions with the Kinematic Sunyaev-Zel’dovich E ffect
Authors: Nick Hand, the Atacama Cosmology Telescope, and the Baryon Oscillation Spectroscopic Survey (58 co-authors)
First Author’s Institution: Department of Astronomy, University of California Berkeley

Today I have the privilege of presenting to you a paper that not only describes the first detection of a cosmic effect that was theorized some forty years ago (as if that wasn’t enough), but was also first-authored by one of Astrobites’s very own: Nick Hand.  Nick first began this work, which describes the first detection of the motion of distant galaxy clusters via the kinematic Sunyaev-Zel’dovich effect, as part of his senior thesis at Princeton University.

OK, so now that I’ve given away the punch line, let’s take a few steps back.  What in the world is the kinematic Sunyaev-Zel’dovich effect? You may have guessed from the word “kinematic” that motion is somehow involved. However, if you faltered at that point, never fear, the name of the effect isn’t meant to be incredibly informative.  Rashid Sunyaev and Yakov Zel’dovich are/were brilliant theorists who in the late sixties/early seventies decided to tackle a specific scientific question: what happens to Cosmic Microwave Background photons when they pass through a massive galaxy cluster on their way to Earth?

The CMB (left) and solar (right) spectra. In comparison, the CMB is MUCH closer to a perfect blackbody.

As you may recall, the Cosmic Microwave Background (CMB) is the remnant of the radiation which was released when the hydrogen in the Universe first became neutral (known as “recombination“) some 13.7 billion years ago (only about 300,000 years after the big bang!). Although initially very hot, this radiation has cooled though cosmic time, such that we now observe it to be a nearly uniform blackbody at a temperature of 2.7 Kelvin.  And when I say nearly uniform, I mean it.  In introductory level astrophysics courses we often like to say the spectrum of the Sun is essentially a blackbody.  However, the CMB puts the Sun – or any other star – to shame (see Figure 1): the Planck satellite measured the CMB to be a uniform blackbody to one part in 100,000.  It really is the closest thing to a perfect blackbody astronomers have ever observed.

However, nothing is truly perfect; there are anisotropies in the CMB (places where it is distinctly not uniform), and these anisotropies can tell us quite a bit about the Universe.  Perhaps the most well known anisotropies in the CMB are those on a relatively large scale (first mapped in detail by WMAP), which are caused by effects such as the size of the horizon at the time of recombination, and matter density fluctuations in the early Universe.  However, these are not the ONLY anisotropies present in the CMB, which finally brings us back to Sunyaev and Zel’dovich.  It turns out that when CMB photons pass through a large galaxy cluster on their way to Earth the result is that the CMB radiation field is distorted in the direction of the cluster.

Figure 2: Example of how the various SZ change the spectra of the CMB. Black Dashed line: the orginalCMB black body, scaled down by a factor of 0.0005. Blue line: the thermal effect. You can see it shifts photons from low to high frequency. Red Line: The kinematic effect (much smaller).

In most contexts when you hear the term “Sunyaev-Zel’dovich effect” (for instance this, this and this astrobite) the authors are specifically referring to the “thermal” Sunyaev-Zel’dovich effect.  In this case, CMB photons are Compton scattered off of hot (hence, thermal) electrons in the center of the galaxy cluster.  The result of this process is to “shift” some of the CMB photons from lower frequencies to higher frequencies, and the effect was first detected in the 1980s.  However, around the same time Sunyaev and Zel’dovich also predicted another effect: if the galaxy cluster were moving (with respect to the rest frame of the CMB) then this bulk motion would also cause scatterings of CMB photons. In this case, the entire blackbody curve is shifted such that the temperature of the CMB appears different in the direction of the cluster. THIS is the kinematic Sunyaev-Zel’dovitch effect, essentially just the Doppler shift of the photon frequency, caused by the motion of the cluster.  The magnitude of the effect is proportional to both the line-of-sight velocity and mass of electrons in the galaxy. Unfortunately, for a large galaxy cluster, the kinematic SZ effect is predicted to be on the order of 20 times weaker than the thermal SZ effect (which is already relatively small – see Figure 2).  Thus, when looking at individual galaxy clusters, scientists have thus far only been able to put upper limits on the distortions due to the kinematic SZ effect, simply due to the difficulty of observing such a weak signal.

So how did the authors get around this?  They did something which is quite simple in concept, but also ingenious (and certainly not simple in application): they stopped trying to look at individual galaxy clusters.   Instead, the authors combined information for literally thousands of clusters by utilizing information from two large survey telescopes: the Baryon Oscillation Spectroscopic Survey (BOSS, part of the Sloan Digital Sky Survey III) and the Atacama Cosmology Telescope (ACT).  BOSS has mapped the three-dimensional locations of  hundreds of thousands of ‘luminous galaxies’,  which are known to often be found in massive galaxy clusters. Thus, the BOSS data set provides the locations of potentially thousands of massive galaxy clusters. The authors have selected the 7,500 brightest of these galaxies that fall within the ACT data region.   They then went to the data from ACT (which maps the CMB at several microwave frequencies), stacked and averaged the data in the direction of the galaxies selected from BOSS in a particular way and were able to detect a non-negligible temperature shift in the region around the brightest galaxy clusters compared to the background.   However, this shift is due to the thermal SZ effect, which is always negative for the frequencies at which the ACT data was taken.  This cannot be done for the kinematic SZ effect, however, because the sign of the distortion depends on the on the sign of the peculiar velocity of the galaxy cluster, and galaxy clusters are just as likely to be moving towards us as away from us.  Hence, when directly averaging together the signal from thousands of galaxies, the signal from the kinematic SZ effect goes to zero.

So what do the authors do instead? One of the real advantages of the kinematic Sunyaev-Zel’dovich effect is that it depends on the peculiar velocity of a galaxy cluster (i.e., it does NOT depend on the velocity of the cluster due to the expansion of the universe), and thus should offer us information on gravitational forces pulling on the clusters (which relates to the formation of structure in the universe). To assess this, the authors computed something referred to as the mean pairwise momentum of their set of 7500 galaxy clusters.  This is essentially a measurement of how much, on average, galaxy clusters are moving either towards or away from one other.  If all of the galaxy clusters have truly random peculiar motions, this value should be zero.  However, from our current theories of structure formation and gravity, we expect massive galaxy clusters to be moving towards one another, on average.

Figure 3: Results showing the motion of galaxies via the kinematic SZ effect (see text)

Figure 3 shows the results from this analysis.  The top panel shows the mean pairwise momentum calculated from this data set (red) as a function of comoving separation (i.e., neglecting the expansion of the universe) of the galaxy clusters, and the expected values based on numerical simulations (bold black line).  The red points clearly deviate from zero, at a statistically significant level of 3.8 sigma.  The bottom panel shows the results for the same analysis, applied to random positions in the ACT map as opposed to the known locations of galaxy clusters. The fact that these points do essentially average to zero gives added credence to the idea that the signal shown in the top panel is real.  Thus, this represents the first detection of the motion of galaxy clusters via the kinematic Sunyaev-Zel’dovich effect.

This discovery truly is a very important first step in opening up a new means of investigation in physical cosmology.  Once it is possible to make more precise measurements of the kinematic S-Z effect, the velocity information it provides will be able to give us additional constraints on the various gravitational forces acting on structures in the universe, such as dark matter and dark energy.

Congratulations, Nick!

If you are interested in a more technical description of the kinematic S-Z effect, check out this link. (Also have a look at the press release put out by Princeton University here).

About Maria Drout

I am currently a graduate student at Harvard University, before which I completed a Masters of Advanced Study in Applied Mathematics and Theoretical Physics at Cambridge University. My main research interests include stellar populations, stellar evolution, and stellar death. I currently work with Alicia Soderberg (Harvard) on observations of supernovae and Phil Massey (Lowell Observatory) testing the current models of massive star evolution. I graduated from the University of Iowa in 2010 where I worked with Cornelia Lang on the magnetized structure of the Galactic Center.


  1. Maria, this is a very nice explanation of the kinematic SZ effect, and congrats to Nick on a great paper!

  2. Maria, why do the authors use momentum rather than velocity?

    Thanks for a great article!

    • Good question, Jeremy. It has to do with the fact the the temperature shift caused by the kinematic SZ effect is proportional to not only the velocity of the galaxy cluster but also to the electron number density. This number density is proportional to the overall mass of the cluster. So, we’re really measuring the momentum (mass x velocity) of the cluster when we measure the kSZ temperature shift.



  1. Voor het eerst direct beweging in clusters van sterrenstelsels waargenomen - [...] dat binnenkort z’n weg zal vinden naar de Physical Review Letters. Bron: Physics World + Astrobites. Gerelateerde …
  2. Dark Flow: A Mysterious Force From Outside of our Local Universe - From Quarks to Quasars - [...] if the galaxy is moving relative to the background glow. This was developed to test this, known as kinematic …

Leave a Reply