Sky & Telescope - July 2021 - 19

rarely merge with similar-size (or bigger) galaxies. As a result,
open clusters cannot escape into the safety of their natal
galaxy's halo, like globular clusters once did.

M4: NOAO / AUR A / NSF

Toward a Complete Picture
These are all interesting ideas, but the complexity of the problem means that we cannot conclusively solve it with pen and
paper alone. The true test must come from running sophisticated computer simulations that combine the formation
and evolution of star clusters with that of galaxies, and then
comparing the results with observations.
In recent years, a variety of teams has started to develop
such simulations, with world-leading groups in the U.S., UK,
and Germany, among many others. This concerted effort of
the astronomy community has brought about a revolution
in globular cluster formation research. For the ﬁrst time, we
have state-of-the-art simulations that can link the conditions of star and cluster formation in the early universe to the
globular clusters that we see today.
I set out with a team of collaborators in Liverpool and
Heidelberg to develop the E-MOSAICS simulations for this
purpose. To date, these simulations are the only ones that
combine a description of all the cluster formation and
destruction physics discussed above (MOSAICS) with a leading galaxy formation model (EAGLE). As such, they are the
ultimate computational testbed for the idea that globular
clusters are the natural outcome of regular cluster formation
in the early universe.
Over the past few years, we have found that the E-MOSAICS simulations do surprisingly well in most areas. They
closely reproduce the properties of young ( " open " ) clusters
forming today, most importantly their numbers of stars and
their ages. Encouragingly, they also simultaneously reproduce
most properties of globular clusters, such as their spatial dis-

Encounters with gas clouds decimate the star
clusters. The biggest clusters in the galaxy
survive the longest.

MESSIER 4
Constellation: Scorpius
Magnitude: 5.4
Age: 12.2 billion years

tributions, their orbital motion around the host galaxy, their
ages and chemical compositions, their large numbers of stars,
and the number of globular clusters per galaxy. In itself, this
is an enormous step, because no theory for globular cluster
formation has ever reproduced the observed properties of the
globular cluster population in such detail. This means that
the premise of the simulations works: Globular clusters can
indeed be seen as the products of regular cluster formation at
early cosmic times.
Now that we have a reasonable model for the origin of

Galaxy mergers throw star clusters out of their
disrupted, gas-rich disks into the surrounding
gas-poor halos.

As the transformed galaxy settles down, the
ejected clusters stay in the halo, largely safe
from disruption. These are the globular clusters
we see today.

s k y a n d t e l e s c o p e . o r g * J U LY 2 0 2 1

http://www.skyandtelescope.org

Sky & Telescope - July 2021

Table of Contents for the Digital Edition of Sky & Telescope - July 2021