The English version of the article that has been published in the ZENIT magazine (January 2021), written by Teymoor Saifollahi and Reynier Peletier.

In the last few decades, the main focus of astronomers in studying galaxies has been to make a consistent picture of the formation of galaxies starting 13 billion years ago and how they evolved to become what we see now. This picture, known as the ‘standard model of galaxy formation’ tells us that the bigger galaxies today, such as our Milky Way, have been formed by the mergers between smaller galaxies. Also, in this picture, more than 80 percent of the matter in galaxies and in the universe is dark matter, a form of matter which is unknown to us since we can not see it.

How do we know about the dark matter in galaxies?

The story of dark matter backs to the 1920s-1930s when astronomers, among them the dutch astronomers Jacobus Kapteyn and Jan Hendrik Oort, mentioned the lack of matter in our galaxy, to describe the observed rotation of the stars in the Milky Way. This subject did not receive much attention until the 1970s when Vera Rubin and Albert Bosma from Groningen among others, studied spiral galaxies other than the Milky Way and found the same issue. In the next decades, astronomers faced the same problem everywhere and now the existing evidence for dark matter is uncountable; in galaxy clusters, large scale structures and cosmic filaments, the cosmic microwave background, everywhere there is strong evidence of sort of unseen matter with different properties than the ordinary matter. In one exceptional case known as the Bullet cluster, it is possible to see that the gravitational effect of dark matter is separated from the observed matter and it is direct evidence that dark matter can not be fixed simply by modifying the equation of gravity.

In time, dark matter has become a necessary ingredient to form galaxies which without it, galaxies could not exist in the way they are now. Nowadays, the question is not whether ”the dark matter exists or not”, it is “what is dark matter?”. To find the answers, nowadays astronomers look at where the dark matter is expected to found the most: dwarf galaxies.

Faint, small and dark

Dwarf galaxies are traditionally defined as galaxies which are at least one order of magnitude fainter, less massive and smaller than the Milky Way. Milky Way as a normal massive galaxy is 100,000 light years across and has 50 billion solar masses of stars while a large dwarf galaxy like the Large Magellanic cloud is 14,000 light years across and has 3 billion solar masses of stars. While dwarfs are faint, they are known to have relatively more dark matter than the stars compare to massive galaxies. A big galaxy like the Milky Way, for 1 gram of stars has 10 grams of dark matter, But for a dwarf galaxy, for 1 gram of stars, it can have 100 to 1000 grams of dark matter. This ratio of dark matter to stars is already explained by the standard model of the galaxy formation.

Few years ago, a group of astronomers reported a population of faint and diffuse galaxies as large as the Milky Way. They also claimed that these galaxies can be as heavy as the Milky Way but with less stars which means they “failed” to convert their mass into stars. Shortly after this claim, astronomers started to measure the masses of these objects. The so-called “ultra-diffuse galaxies” or “UDGs” do not fit well within the traditional definition of dwarf galaxies because they are faint and large. Despite that these objects are known for a few decades, because there has not been a reliable distance and size measurement of these objects, they had not received the attention that they deserved.

The total mass of a galaxy including stars, gas and dark matter is estimated by measuring the motion of stars within a galaxy. In spiral galaxies, stars tend to be on a disk which rotates around the centre of the galaxy. This rotation is the key to estimate the mass of the galaxy in different radial distances from the centre. However, not all the galaxies show such a clear sign of rotation. In fact, for most of the galaxies, such as ellipticals, stars move randomly and measuring this randomness can be translated to a mass and it is not possible to measure the enclosed mass in different radial distances. Also, when an object becomes fainter, the observing time increases and that is why doing such measurements for UDGs is not a simple task.

Investigations to measure mass of the UDGs so far studied a handful of UDGs. Among the studied UDGs, the UDGs in the Virgo cluster, the closest galaxy cluster to us at the distance of 50 million light years, have received more attention. VCC1287 was the first to be studied. This galaxy has a mass of about 100 billion solar masses, which is about 10 times less than the Milky Way and is the same order as the Large Magellanic cloud. Later, 3 other UDGs known as VLSB-B, VLSB-D and VCC615 were studied and interestingly, each was very different from the other. Soon, all these studies concluded that while UDGs have 100 – 1000 times more dark matter than their stars, as expected for dwarf galaxies, they are not as heavy as the Milky Way.

However, up until now, there was one exceptional case: DF44.

Astronomers being creative: tricks to measure the mass faster

There are different ways to measure the combined mass of a galaxy. Doing spectroscopy and measuring the kinematics of stars in galaxies is a common method but it is not the only one. It is known that galaxies are surrounded by several globular clusters, dense accumulation of a few hundred thousand stars, which move randomly around the galaxy. In time, astronomers found that more massive galaxies have more globular clusters and less massive galaxies have less and they can use the number of globular clusters to estimate the mass of the galaxy, stars, gas and dark matter combined.

Later, it was proposed that instead of doing long and expensive spectroscopic observations to measure the mass, it is possible to take deep images of galaxies which need 1000 times shorter observing time times than spectroscopy and count the number of globular clusters around galaxies. This method is purely observational and astronomers do not have a complete model to explain why such a relationship between mass and number of globular clusters exist. It has been shown that this relation can be simply a statistical consequence of the hierarchical galaxy formation which says that smaller galaxies merge to make the bigger galaxies.

In a simplified picture, we can assume that galaxies are formed from several “galaxy seeds” with the same mass. Therefore, larger galaxies are made of more galaxy seeds and smaller galaxies are made of less galaxy seeds. For some reason, every seed gives birth to more or less the same number of globular clusters. This means that at the end, depending on how many galaxy seeds a galaxy collected, it also collected the proportional amount of globular clusters.

The simplified scenario here explains that some part of this relationship between mass and number of globular clusters can be statistical, regardless of the formation scenario of the globular cluster itself. In general, globular clusters look simple, but we do not know how exactly they formed. They are among the oldest objects in the universe and once a while, a new study finds something fascinating about them. Unfortunately, we can only study the Milky Way globular clusters in detail since they are resolved to their stars and globular clusters around other galaxies look like a point to us. However the upcoming surveys and telescope will make astronomers able to study them in a way that has not been possible before.

The story of DF44

Dragonfly 44 or DF44 is a large UDG that is located in the Coma galaxy cluster at a distance of about 330 million light-years. This galaxy was one the first to be discovered, among the other UDGs that were discovered by Pieter van Dokkum and the Dragonfly team. The large size of this object got the attention of this group and that is the main motivation that they chose this galaxy to study in detail: If any of the UDGs are supposed to be weird, it most likely to be the largest one.

In 2016, they measured the mass of the galaxy with ordinary 2D spectroscopy and found that this galaxy is as massive as the Milky Way. Since this galaxy has 10,000 times fewer stars than the Milky Way, it means that it is 99.99% dark matter. In other words, for 1 gram of stars, it has 10,000 grams of dark matter. This is at least 10 times more than the dwarf galaxies that have 100-1000 times more dark matter than stars. In their study, they also used deep images of the galaxy and counted about 100 globular clusters around the galaxy. With this number of globular clusters, this galaxy will be massive. Therefore, two different methods to measure the mass, spectroscopy and globular cluster count, independently showed that the galaxy is massive.

The globular cluster count in 2016 was based on ground-based observations. These observations suffer from the blurring effect of the atmosphere and because of that, the images look blurred. That is why, van Dokkum and colleagues made new observations in 2017 and this time, they used the Hubble space telescope (HST) to make high-resolution images of the galaxy. Their study in 2017 led to counting ~80 globular clusters around DF44, a number which obviously is slightly smaller than before, but still in the same order.

In 2019, they did 3D spectroscopy with the Keck 10m telescope in Maunakea in Hawaii and re-calculated the mass. Interestingly, the mass that they found was smaller than before and was very similar to dwarf galaxies. However, the mass measurements were very incertain. It means that they reported a range of possible masses for the galaxy and in this range, it is hard to say if the galaxy is massive or dwarf. The 80 stars clusters from the 2017 study was their proof to say that the galaxy is massive.

Why 99.99% dark matter is a problem?

Dark matter is fundamental for the universe to be, it is the main ingredient of the galaxies and without dark matter galaxies do not exist in the way they are now. In the standard model of galaxy formation, 13 Billion years ago and in the beginning of the universe, dark matter particles started to accumulate on top of each other and made dark matter halos. Later, the gas in the universe which is the main fuel for star formation, was attracted to the dark matter halos and formed stars. This mixture of dark matter, gas and stars were the proto-galaxies. In the proto-galaxies, more massive dark matter halo means attracting more gas and making more stars but the ratio of the dark matter to stars is expected to remain the same. On the other hand, larger and more massive halos have stronger gravity and are better in keeping their gas and less massive halos are not. That is why, in time, smaller halos lost part of their gas for different reasons such as the strong winds from supernova explosion in the galaxy and failed to make more stars. So eventually, the smaller halos, that turned to be dwarf galaxies in the current time, ended up with relatively more dark matter than their stars compared to more massive halos.

In this recipe, a dwarf galaxy can have 100-1000 times more dark matter than stars. But the case of DF44 with 10,000 times more dark matter than stars is not possible. If this galaxy is indeed as massive as the Milky Way, how did it fail to attract the gas and make stars as other massive galaxies did? That is why the hypothetical galaxies like DF44 are given the notion “Failed Milky Way”.

Problem is solved: no dark and big galaxy after all.

In the last few years DF44 has been an anomalous galaxy and a problem for the standard model of galaxy formation. Since it was the only object with this behaviour, it motivated us to have another look at the data and re-calculate measurements. In the beginning, we did not know that the result would be that different at the end and it was an attempt to reproduce the previous studies. For the case of DF44, the claimed number of globular clusters was the main evidence for being a massive galaxy. Therefore, we started to re-analyze the HST images of the 2017 study of Pieter van Dokkum and soon, after a few weeks, it became clear to us that the result would be different.

We started to do a more careful analysis for detecting globular clusters since we expected that to be the problem. Later on, through the analysis we found that, even though our method in the globular cluster detection does not change the outcome directly, we could detect that globular clusters around DF44 are more concentrated than what the 2017 study assumed. In other words, the 2017 study has an inaccurate assumption about how globular clusters are spread around the galaxy. What the team of Pieter van Dokkum has done is a bit contradictory itself: they assume something about DF44 based on what we know about dwarf galaxies and then at the end find out that the galaxy is not a dwarf galaxy.

In our recent research at the Kapteyn Astronomical Institute of the University of Groningen (Netherlands), with the participation of the Instituto de Astrofísica de Canarias (IAC) and within our european collaboration known as SUNDIAL, we re-counted the number of globular clusters around DF44 and found 20. This is four times less than the previous work which found 80. This simply means that this galaxy has dark matter between 100-1000 times its stars, as other known dwarf galaxies. In this work, we did try all the possible cases with any other assumption and they all led us to this number.

What’s next?

In science, assumptions have to be made and sometimes mistakes are made. Our result from this work does show that DF44 is not extraordinary and probably there is no massive and dark galaxy. However, in the last few years, there have been few controversial observations of dwarf galaxies and UDGs. Objects such as NGC1052-DF2 and NGC1052-DF4 with an intense debate on their mass and distance. All the dwarfs and UDGs so far have been observed in galaxy groups and galaxy clusters, where 100 to 1000 of galaxies are located close to each other. Recently there have been few observations that found UDGs environments outside of galaxy groups and galaxy clusters. These UDGs look to have less dark matter than expected. All these new findings is because astronomers started to study faint objects in a way that was not possible before. It takes a bit of time to observe many objects and collect the necessary data to achieve a conclusion on the dwarf galaxies. Certainly, there will be more surprises in the near future and it is just the beginning.

For DF44 it already has been ended.