Australasian Science Magazine

An image of the centre of the Coma Cluster. Green dots show the distribution of thousands of faint dwarf galaxies, which do not include the recent discoveries of large, ultra-diffuse galaxies. In this crowded region of space, galaxies will frequently interact gravitationally with one another and the underlying dark matter of the cluster itself. Credit: NASA/JPL-Caltech/GSFC/SDSS

By Cameron Yozin

A recent discovery in the Coma Cluster highlights the important role of dark matter in transforming galaxies.

The billions of galaxies that make up the known universe come in a dazzling array of masses, shapes and colours. In the past few decades we’ve made incredible advances when it comes to observing these galaxies and better understanding how such diversity arose in the universe, yet our observations continue to surprise us.

To understand why galaxies look different, consider the fact that people, born in different parts of the world and going about their individual lives while interacting with different people, look different. Similarly, we find that the appearance of galaxies, ranging from the smallest with only a few thousand stars to the largest with more than a trillion times the mass of our Sun, are a product of the environment in which they form and the interactions they then have with other galaxies.

While people have some choice over the direction their lives take, the fate of galaxies is controlled by the all-permeating force of gravitational attraction. In many cases, this causes them to collide in spectacular fashion, leading to the formation of yet more massive galaxies. In other cases, a galaxy may be drawn towards the combined gravity of many other galaxies lying in a single gravitationally-bound structure known as a galaxy cluster.

Transforming Galaxies

The Coma Cluster,which lies 300 million light years from us, is a relatively close and massive example of a cluster of thousands of galaxies that have been trapped in a cosmic dance for billions of years. Most of these galaxies could be described as “dead”: their elliptical shape and red colour implies that they have formed no new stars in a long time.

In contrast our home galaxy, the Milky Way, is a “living” spiral disk emitting a relatively blue colour thanks mostly to ongoing star formation. According to the prevailing theory of galaxy formation, our Milky Way more closely resembles the disk shape that most galaxies are thought to have been born with.

This disparity in colour and shape appears to be directly related to the different environments in which these galaxies exist. Whereas the Milky Way has so far resided in a relatively low density region of the universe, on the outskirts of the Virgo Cluster, the Coma Cluster has a high density of galaxies, making it a dangerous place for galaxies wishing to retain a disk shape and continue to grow through the formation of new stars.

Close gravitational interactions between galaxies can “heat” or distort their disks so that they thicken to a more rounded shape or are destroyed entirely. At the same time, the cold hydrogen gas within a galaxy, required to fuel star formation, can be lost when the galaxy is drawn into a cluster. In much the same way as wind pressure pushes against your hand if it’s held outside a moving car, this cold gas is pushed out of the galaxy by hot dense gas lying inside the cluster.

These same methods of transforming galaxies have been observed among the roughly 50 known small galaxies that orbit our Milky Way. As these “dwarf” galaxies are close enough to us for detailed analysis of their stellar populations, we have discovered that they were “killed” almost as soon as the gravitational attraction of the Milky Way took hold of them as much as 10 billion years ago.

The singular exceptions are the blue, disk-shaped Magellanic Clouds. Visible as fuzzy patches of light on a clear night in the Southern Hemisphere, the Magellanic Clouds are believed to have only started to orbit the Milky Way in the past few billion years.

An Unexpected Discovery

The mechanics of galaxy transformation aren't fully understood yet. For example, we don't know exactly how long it takes galaxies of different masses to be affected by their respective environments.

While it was generally accepted that disk galaxies do not survive long in clusters, discoveries since late 2014 have not only contradicted this belief – they have revealed some of the most extreme disk-shaped galaxies ever found.

The story starts when a North American team, using an innovative new telescope comprised of camera lenses designed for wildlife photography, turned their gaze to the Coma Cluster. Originally they were looking for faint light in between galaxies as evidence of stars removed by gravitational interactions. Instead, they found galaxies so faint that earlier observations of the cluster made with higher resolution telescopes most likely looked straight through them!

Analysis of these new galaxies produced some startling results (Fig. 1). Some were as large in size as our Milky Way but with only 1% of their mass tied up in stars. With starlight spread so thinly throughout this new type of galaxy, they became known as “ultra-diffuse” galaxies. Their discovery motivated astronomers to revisit the archived observations of the Japan’s Subaru telescope, and the found that there could be thousands of these galaxies located throughout the Coma Cluster. Then, when another team used the same technique to look at Virgo, a near-neighbour of the Milky Way, they not only found more ultra-diffuse galaxies but galaxies that were even larger and even more diffuse.

No galaxies with such extreme properties had previously been detected, especially in a dense cluster environment. The closest type of galaxies with properties similar to the new ultra-diffuse galaxies are known as “low surface brightness” galaxies, but both galaxy formation theory and actual observations consistently show that these galaxies cannot exist in dense regions like clusters. In the case of the ultra-diffuse galaxies, their red colour and widespread distribution in both the low and high density regions of the Coma Cluster implies that they were gravitationally drawn into it and stopped forming stars as early as seven billion years ago, which is half the age of the universe.

The Role of Dark Matter

How do we explain these puzzling observations? Since these galaxies have only recently been discovered, astronomers have only limited data with which to find a solution. What we do know, however, is that dark matter must play an important role.

Dark matter is the elusive substance that makes up about 80% of all matter in the universe. It doesn’t interact with electromagnetic waves (light), and is therefore invisible to us. However, there is very strong evidence that it exists, starting with the discovery in the late 1970s that the speed of stars within many galaxies could not be possible if there was not some additional invisible mass binding them together.

Further evidence lies in the distortion of the path of light when it passes near massive objects, as predicted by Einstein's theory of general relativity, and observed directly in cases such as the Coma Cluster. In fact, it is by measuring this distortion that the total mass of the Coma Cluster can be found, revealing not only that it is 1000 trillion times as massive as our Sun, but that only about 10% of this mass is in a visible form such as stars or gas.

Just as gravitational interactions between galaxies in the Coma Cluster can transform them, so too can an interaction with the Coma Cluster's substantial dark matter. Most of this mass is concentrated at the centre of the cluster. Therefore, any ultra-diffuse galaxies close to this centre are most at risk of having their disk shape destroyed.

One method by which a galaxy can slow down the process of being destroyed by the gravity of another object is by being relatively massive itself. This makes it possible to perform a simple calculation to work out how much mass an ultra-diffuse galaxy requires to avoid being dramatically transformed through interaction with the dark matter in the Coma Cluster. The result: some of these galaxies must have had 100 to 1000 times as much mass tied up in dark matter than in stars when they fell into the Coma Cluster.

But how can more than 99% of a galaxy’s mass be dark matter if the average amount of dark matter in the universe is only 80%? In the case of the ultra-diffuse galaxies, the mass of non-dark matter material could be extremely low if cold hydrogen gas had been removed from the galaxies before they could form most of the stars they would have if the galaxies had not fallen into the Coma Cluster (Fig. 2).

Figure 2. An artist’s impression of how a galaxy can be transformed from a normal blue star-forming disk galaxy to an ultra-diffuse object when drawn into the Coma Cluster by its immense gravitational attraction. Credit: C Yozin, ICRAR

Simulating the Death of Galaxies

This solution to the puzzle fits nicely, but we need to test the idea more rigorously to see if it works. Astronomy is unique among sciences insofar as it is almost impossible to do experiments with real galaxies, so the best option is to build a detailed model of a galaxy and its environment inside a supercomputer. By incorporating into our model all the physics and chemistry we believe play an important role in the evolution of the galaxy, we can simulate its evolution over billions of years and check its properties to see if they match those of the galaxies we see in our observations.

This method was applied to the case of ultra-diffuse galaxies to ask two key questions: can the hot gas of the Coma Cluster push out the cold gas of a galaxy that fell into it as early as seven billion years ago? And can a disk-shape satellite of Coma survive for this long if it contains 99% dark matter?

To answer both questions, we need to know the mass and composition of the Coma Cluster in the past. Unfortunately, we have no time machine to go back in time to find this out, so instead we recall the fact that looking deeper into space is akin to looking back in time, given that the speed of light is finite.

So, if we can find real galaxy clusters in the distant past that resemble an earlier version of the Coma Cluster, we can base our ultra-diffuse galaxy cluster model on it. To work out what an earlier version of the Coma Cluster should look like, we refer to the results of recent supercomputer simulations of the universe, which have shown in great detail how Coma-sized collections of galaxies have been assembled over cosmic time.

The next step is to build a model of a galaxy that, when put into our model of the Coma Cluster and allowed to evolve over billions of years, ultimately resembles the ultra-diffuse galaxies that we’ve observed. To do this, we refer to the theory of galaxy formation that explains how various properties of a primitive galaxy, such as its size and gas content, are correlated with its dark matter mass.

By running these simulations on supercomputers located at The University of Western Australia, we can successfully reproduce the theorised scenario for the formation of the ultra-diffuse galaxies and see that, despite such an immense amount of dark matter, these galaxies cannot survive indefinitely. The dark matter in the Coma Cluster continuously steals the dark matter and stars of the galaxies orbiting within it, gradually weakening their resistance and leading to their eventual demise.

The Next Step

Although these simulations were successful, our work in understanding this new type of galaxy has only just begun. By supporting the idea that a dark matter dominated galaxy can survive the chaotic environment of the Coma Cluster, future observations with state-of-the-art telescopes will be undertaken to find indirect evidence of this mysterious substance, using the individual motions of stars in the ultra-diffuse galaxies. We can also establish how long these galaxies have resided in the cluster by measuring their velocity distribution.

These future discoveries will allow us to refine or revise our ideas about the origin of this unusual type of galaxy, leading to improved models that may answer some significant questions.

• Do these galaxies exist only in clusters?
• Did these galaxies fall into a smaller cluster first, before that cluster fell into the Coma Cluster?
• Have these galaxies actually fallen into the Coma Cluster only recently, to be observed shortly before their destruction commences?
• Did these galaxies stop forming stars through some other means, such as when their most massive stars exhaust their fuel and explode with enough energy to blow out all the cold hydrogen gas initially contained in the galaxy?

With such extreme properties, the exciting discovery of these ultra-diffuse galaxies will aid in our understanding of the extra­ordinary diversity in galaxies and the processes involved in their evolution.

Cameron Yozin is a research assistant at The University of Western Australia node of the International Centre for Radio Astronomy Research.