The National Institute for Computational Sciences

Mysteries of the Milky Way

Reconstructing Our Galaxy's Formation History

By Hanneke Weitering


For billions of years, galaxies have been fighting the ultimate star wars, attacking one another via head-on collisions, harassment, and even cannibalism. The interactions can involve mergers, disfiguring high-speed flybys, or even the ingestion of dwarf galaxies.

The transformative relationships between galaxies cause them to be of different shapes and sizes: spiral, round, elliptical, or even amorphous from having been disrupted.

In photographs of deep space, galaxies might look like randomly assorted static blobs and swirls of stars and dust, but in reality these blobs are highly evolved, complex entities whose structures help reveal details about the past.

Even though we are in the Milky Way, we know far less about it than we do other galaxies. The reason is that we can easily observe other galaxies from the sky as outsiders, but our location inside the disk of the Milky Way makes seeing the whole picture impossible.

Observations through the plane of our galaxy's main disk as well as examinations of other spiral galaxies led to the assumption that our galaxy is also a spiral, but no one has ever actually seen it from an external vantage point to be sure.

Probing Our Galaxy's History

Although most astronomers believe the Milky Way is spiral shaped, they are not all in agreement as to the details of its structure, much less how it evolved. However, scientists now may have found a way to unravel the long-standing and elusive mystery behind our galaxy's formation history.

Researchers at Michigan State University (MSU), the Massachusetts Institute of Technology (MIT), Duke University, and the Space Telescope Science Institute (STScI) in Baltimore are devising a method for reconstructing the evolution of the Milky Way and its orbiting dwarf galaxies.

New stellar surveys such as the Sloan Digital Sky Survey and SkyMapper have helped astronomers build 3D maps containing everything from asteroids inside the solar system to quasars located billions of light-years away. Using advanced statistical tools, the researchers compare these newly available observations with their own theoretical models to get a better idea of how our galaxy may have been formed.

"Our hope is that with this analysis we will be able to isolate a subset of formation histories that can best reproduce the properties observed," explains Facundo Gómez of MSU. "In other words, we would like to put constraints on the formation history of the galaxy."

Spanning roughly 100,000 light-years across and containing an estimated 300 billion stars, the Milky Way is not exactly easy to model, and simulating its entire history is substantially more difficult. With the help of supercomputing, however, the task becomes much more feasible.



Principal investigators Gómez, Brian O'Shea of MSU, and Christopher Coleman-Smith of Duke were granted 950,294 compute hours on the Kraken XT5 supercomputer at the National Institute for Computational Sciences (NICS) for this project. With the allocation they were able to generate 12 simulations of different possible formation histories. The researchers used the Nautilus supercomputer, also managed by NICS, to create the initial conditions of the simulations.

To study the formation history of the Milky Way, scientists must look far beyond the spiral disk to the stellar halo—a faint, spherical population of stars that encompasses the entire galaxy. Inside this dim shell of stars lie the remnants of all the satellite dwarf galaxies that the Milky Way has ever devoured.

Over time, the stars from all these cannibalized dwarf galaxies have become mixed up with their fellow halo stars. The key to solving the puzzle of what stars came from what dwarf lies within the chemical composition of the stars.

Because all galaxies have their own evolutionary histories, we can distinguish their stars based on their distinct chemical composition. For instance, larger satellite galaxies are typically denser and more chemically enriched. The chemical enrichment is a result of multiple generations of stars fusing hydrogen and helium into heavier elements and subsequently spewing them out into the interstellar medium when they explode as supernovae.

Smaller galaxies have less gravity, and so they also retain less "star stuff" from their supernovae, leaving them less dense and less chemically enriched than their larger counterparts. Determining the elements that make up the stars can be done observationally using spectroscopy. Knowing exactly what makes up each of the halo stars is key in figuring out which stars came from which ancient dwarf galaxies.

Another clue about the Milky Way's cannibalistic history lies in the locations and motions of the stars within the halo. "In the models of the formation of stellar halos we find that their inner regions typically formed first and hence contain information about the most ancient accretion events that a galaxy has experienced," Gómez says. "As a consequence, it is important to sample large regions of the galactic stellar halo, as we can get information about the formation history of the galaxy at different times in its evolution."

In addition, Gómez explains that stars from the same satellite galaxies tend to have similar velocities. Measuring the velocities of halo stars can also help with distinguishing their origins.

The Gaia Project

Conveniently, the European Space Agency launched a satellite just last year to do exactly that—measure the positions and velocities of approximately one billion stars. The Gaia project will serve as a census of the Milky Way and its surroundings, providing data that is crucial to understanding the evolution of the Milky Way.

"The Gaia survey will allow us to group stars that came from different satellites," Gómez says. He and his team are on the edges of their seats waiting to see the data and compare it with their theoretical models.

Simulations and Observations

"A very exciting part of my research is to compare the results from our chemo-dynamical simulations to observations of the real galaxy," Gómez says. He explains that even with their most accurately reproduced simulations, there are always some unexpected observables that they cannot match. "This is very exciting because these disagreements teach us about the physics we are missing and show us the way to further improve our models."

Although this team of galactic archaeologists focuses on reconstructing the Milky Way's stellar halo, the tools and methods they have devised will be useful for a variety of other future projects that go beyond the realm of astronomy.

"The statistical tools that we have developed can be applied to any problem that involves models with large numbers of input parameters," Gómez explains. Their software is available for anyone to use. See the Related Links below to the Models and Data Analysis Initiative (MADAI) to download the MADAI Distribution Sampling tools.

[Two discrete research groups that were engaged in probing the formation history of the Milky Way galaxy ultimately joined forces for increased overall efficiency and greater scientific impact, and the project pursued by the two groups in collaboration has been the subject of this science feature article. In addition to principal investigators Gómez, O'Shea, and Coleman-Smith, other members of the original research group were Robert Wolpert of Duke and Jason Tumlinson of STScI. Joining the project when the group expansion took place were Brendan Griffen and Anna Frebel of MIT.]

Related Links

Article posting date: 20 May 2014

About JICS and NICS: The Joint Institute for Computational Sciences (JICS) was established by the University of Tennessee and Oak Ridge National Laboratory (ORNL) to advance scientific discovery and state-of-the-art engineering, and to further knowledge of computational modeling and simulation. JICS realizes its vision by taking full advantage of petascale-and-beyond computers housed at ORNL and by educating a new generation of scientists and engineers well versed in the application of computational modeling and simulation for solving the most challenging scientific and engineering problems. JICS runs the National Institute for Computational Sciences (NICS), which had the distinction of deploying and managing the Kraken supercomputer. NICS is a leading academic supercomputing center and a major partner in the National Science Foundation's eXtreme Science and Engineering Discovery Environment, known as XSEDE. In November 2012, JICS sited the Beacon system, which set a record for power efficiency and captured the number one position on the Green500 list of the most energy-efficient computers.