Monthly Archives: September 2016

Supernovae Destroy Dwarf Galaxies: Dark Matter is Safe

The existence of dark matter has not exactly been under threat – the ratio of dark matter to ordinary matter in the universe is well established, at about 5:1 in favor of dark matter. Consistent results are found between observations of the cosmic microwave background, observations of clusters of galaxies, and observations of the rotation curves of galaxies. (The MOND theory as an alternative to dark matter does not do well at scales greater than that of individual galaxy rotation curves.)

But there has been an issue around galaxy formation. It has been expected that many more dwarf galaxies should be seen in our Local Group, which is dominated by the Andromeda Galaxy (#1) and our Milky Way Galaxy (#2, sorry folks), along with the aptly named Triangulum Galaxy (#3).

Where are the Dwarfs?

Our Milky Way has only around 30 dwarf galaxies as companions, the best known of which are the Large and Small Magellanic Clouds. While a few more have been discovered only recently, simulations of galaxy formation have previously suggested this number ought to be more than 1000! This posed a problem for both our understanding of dark matter and our understanding of galaxy formation.

Now, from CalTech comes a much more detailed simulation of how galaxies similar to the Milky Way are formed. The researchers used over 700,000 CPU hours of supercomputer time to create the most detailed simulation ever of the galaxy formation and evolution processes.

“In a galaxy, you have 100 billion stars, all pulling on each other, not to mention other components we don’t see like dark matter. To simulate this, we give a supercomputer equations describing those interactions and then let it crank through those equations repeatedly and see what comes out at the end.”  – Caltech’s Phil Hopkins, associate professor of theoretical astrophysics.

Death by Supernova

Postdoc Andrew Wetzel and Prof. Hopkins paid special attention to the effects of supernovae. When supernovae explode they release tremendous amounts of kinetic energy. They generate powerful winds that reach speeds of over a thousand kilometers per second.

In a dwarf galaxy an individual supernova can have substantial effect. The researchers’ simulations indicate that dwarf galaxies can actually be destroyed by the effect of even a single supernova during their early history. Stars and gas that would form future stars can both be blown out of the dwarf galaxies. In addition, many dwarf galaxies in the Milky Way’s neighborhood would have been destroyed by the gravitational tidal forces of the Milky Way, the simulations show.

These advanced galaxy evolution simulations appear to solve the dark matter and dwarf galaxy problem. The authors plan to refine their results and develop even greater understanding of galaxy formation with simulations of even greater power in the future.


Simulated View of Milky Way Galaxy
The formation and evolution of the galaxy were done on a supercomputer. Credit: Hopkins Research Group/Caltech



Dark Matter Clumps Tear up Clusters

What is destroying globular clusters?

Globular clusters were formed early in the history of our Milky Way’s history; the 150 or so globular clusters in our galaxy contain many of its oldest stars. Globular clusters are round (hence their name), dense, gravitationally bound collections of stars and can contain hundreds of thousands of stars.

What’s older than globular clusters? Dark matter subhalos! Our galaxy is dominated by dark matter distributed in a halo. Massive supercomputer simulations have shown that regions of higher dark matter density known as subhalos were the seeds for the formation of the galaxy. These subhalos, with millions of solar masses, formed first and supplied the gravity necessary for galaxies to subsequently begin their formation.

Palomar 5 is smaller than most globulars. It was detected only in 1950, in part due to its low mass of only 16,000 solar masses. Palomar 5 is far above the Milky Way’s disk, residing in the dark matter dominated halo, and has been heavily influenced tidally over the past 11 billion years. In its next encounter with the disk, some 100 million plus years into the future, it may even be  completely torn apart by tidal interactions.

Palomar 5 shows significant tidal disruption, with a very long stream of stars trailing out of the cluster, pulled out by tidal forces. The length of the stream is several tens of degrees across the sky, some 30,000 light-years in extent. This is greater than the distance from the Sun to the center of the Milky Way. The stream’s mass is 5000 times that of the Sun.

Of great significance are two well defined gaps in the stream. These gaps are very intriguing to astrophysicists, because they may be probes of the nature of the dark matter in our galaxy’s halo.


Upper portion of Figure 9 from Erkal et al. (referenced below). The two gaps are centered on the dotted lines.

Recently three astrophysicists from the Institute of Astronomy at Cambridge University have modeled the stream and these two gaps and described three main possible causes of the gaps: the Milky Way’s bar (our galaxy has spiral arms leading into a central bar), giant molecular clouds, and/or dark matter halos. (Erkal et al. paper in References below).

They find that gravitational interaction from giant molecular clouds might explain the smaller gap, but not the larger one. Interaction from the Milky Way’s bar is another possibility, but might not be the best for producing such clean gaps, that on the face of it, seem to be due to discrete encounters with smaller structures.

Because of the well defined nature of the gaps, the researchers’ preliminary conclusion is that dark matter halos caused both, and especially so in the case of the larger gap. The smaller gap could be caused by giant molecular clouds, but probably not the larger gap.

The leading tail of the star stream (shown on the left side of the figure) has a two degree gap; this is consistent with an interaction from a dark matter subhalo of 1 to 10 million solar masses. The trailing tail has a nine degree gap that is consistent with perturbation of the stream due to a dark matter subhalo of 10 to 100 million solar masses.

Additional data from several planned experiments should allow better discrimination between the possible causes of the gaps. It is very interesting to note that if the smaller gap is due to a sub halo of a few million solar masses, that knowledge in turn can be used to constrain the mass of the dark matter particles to be greater than 2% of the electron rest mass. This would rule out axions as the dominant contributor to dark matter; the axion mass is expected to be much less than 1 electron-Volt (eV) whereas the electron mass is 511,000 eV.


Erkal, D., Koposov S., and Belokurov V. 2016 “A sharper view of Pal 5’s tails”

Kupper, A. et al. 2015 “Globular Cluster Streams as Galactic High-Precision Scales”

Kuzma, P. et al. 2014 “Palomar 5 and its Tidal Tails”