High-Res Simulations Help Solve Missing Massive Satellites Problem
by Alyson Brooks | Grainger Postdoctoral Fellow, Astronomy Department UW-Madison
Posted Sep 18, 2012
Via Lactea computer simulation of a Milky Way type galaxy, showing thousands of smaller satellite galaxies. Credit: Jurg Diemand and the Via Lactea Project, http://www.ucolick.org/~diemand/vl/
The Pan-Andromeda Archaeological Survey (PAndAS) is studying our neighbor to find faint dwarf galaxies and learn about them. Credit: Geraint Lewis
“The ordinary matter that we are made of and that we interact with in our daily lives is just a small amount of the total matter in the Universe. The Universe is mostly made up of cold dark matter, a mysterious form of mass that only interacts with us through gravity,” explains Alyson Brooks, the Astronomy Department’s first Grainger Postdoctoral Fellow.
“The existence of cold dark matter does an excellent job of reproducing the distribution of galaxies in our Universe, and explains things like why our Milky Way isn't flying apart, given how fast it's spinning. (All that extra gravity from cold dark matter holds us together.) But cold dark matter has had a very rough time explaining observations on the smallest scales—the centers of galaxies and the smallest galaxies,” she continues.
The most famous problem with cold dark matter, the “missing satellites problem,” was first highlighted in 1999. When theoretical astrophysicists run large-scale simulations of galaxies like our own Milky Way, they find that there should be hundreds to thousands of smaller "satellite" galaxies in orbit around us. However, we observe only dozens of these galaxies. “Where are the remaining satellites?” Brooks asks. “A related problem is that the most massive satellites predicted by the simulations are always more massive than any we see around our Milky Way or our nearest neighbor, the Andromeda galaxy. Where are the missing, massive satellites?”
The highest resolution simulations used to study the predicted population of satellites are usually run without gas and stars, and include only cold dark matter. This is because simulations that include gas and stars are much more computationally expensive at a given resolution than dark matter only, and high resolution is required to study the tiniest satellites. The highest resolution runs that include only dark matter already take millions of CPU hours (months of time) to run on supercomputers. Because there's so much more mass in dark matter than normal matter, astronomers have assumed that the gravity of the dark matter will cause the normal matter to follow the structure of the dark matter.
But as computational power increases, theorists are now entering a phase where they can run simulations with gas and stars at very high resolution. Brooks is part of a team producing these high-resolution simulations. What they find when they include gas and stars dramatically alters the expectations of what we should see in satellite galaxies around the Milky Way.
In a series of papers to be published later this year, Brooks and her collaborators have shown that the inclusion of stars that go supernovae can alter the dark matter structure of galaxies, quite literally pushing dark matter out of the centers of satellites. The overall effect is to lower the masses of the bright satellite galaxies we see. “When the masses predicted by dark matter-only simulations are lowered by supernovae, the predicted and observed masses of the Milky Way's satellites are brought into agreement, solving the missing, massive satellite problem that has haunted astronomers for more than a decade,” says Brooks.
Brooks and her team are now expanding the work, demonstrating how the inclusion of normal matter impacts the lifetime of satellites. When gas is included in simulations, it forms a disk at the center of galaxies, like the gas and stars that make up the disk of our Milky Way. Dark matter alone does not form a disk galaxy. The presence of the disk can shred satellite galaxies orbiting within a Milky Way galaxy, altering the predicted number of satellites that should survive to the present day.
The NASA High-End Computing (HEC) Program through the NASA Advanced Supercomputing (NAS) Division at Ames Research Center provided resources supporting this work, and the National Science Foundation and The Grainger Foundation supported this research.