A Quest to
Understand Infinity

Astronomers use XSEDE/TACC resource Stampede to map the Andromeda galaxy

Gazing up at the stars on a clear night often evokes a desire to contemplate the unknown. Past our view, through Earth’s atmosphere, and into deep space lie galaxies that stretch across cosmic time and into infinity. And inside these galaxies are stars that hold the secrets to the history of our existence.

Hubble SpacecraftHubble Telescope

Since the beginning of humanity, astronomers have puzzled over worlds beyond our own, and have endeavored to map this unexplored territory. But charting space is a daunting task, and making these observations are efforts that continue today.

A group of astronomers from around the world is rising to the challenge of unlocking some of the universe’s profound mysteries. They began a four-year program in 2011, the Panchromatic Hubble Andromeda Treasury (PHAT), which represents the most ambitious mapping project completed with the Hubble Space Telescope.

“This project was nearly the largest amount of Hubble Space Telescope time ever awarded,” Julianne Dalcanton, principal investigator of the PHAT project said. “Our group is kind of like the Fast and Furious team, we have over 30 astronomers, both senior and junior levels, from around the world.”

Using resources from the National Science Foundation Extreme Science and Engineering Discovery Environment (NSF XSEDE) and the Texas Advanced Computing Center (TACC), the group is mapping and studying millions of individual stars that comprise the Andromeda (M31) galaxy. Their research led to the most detailed, high-resolution panoramic image of the galaxy and numerous publications— the most recent on M31’s stellar initial mass function was published this February in the Astrophysics Journal.

More than 2,000,000 million light years away from Earth is Andromeda, a rotating disk of stars, gas, and dust. Andromeda is part of the Local Group, a cluster of 80 galaxies including our own, the Milky Way.

“It’s hard to study a galaxy when you live inside of it. But it turns out that Andromeda, a neighbor of ours, is a lot like the Milky Way in terms of its size and chemical composition,” said Daniel Weisz, an astronomer in the PHAT group. “By studying how Andromeda formed and evolved, we’re actually getting clues to how our own galaxy has formed and changed over time.”

The Hubble Space Telescope (HST) is the tool that allows the team to capture and record vivid details about the galaxy. Because the telescope orbits Earth, it can provide information that ground-based telescopes cannot.

Spiral arms:


Any elongated and curved spiral sections that are connected to the center of a spiral galaxy.

“Typically in astronomy when you look at a pretty picture of famous galaxies, you see spiral arms. Those are made up of many stars — it’s just that the galaxies are so far away, you can’t see each star,” Weisz said. “With the angular resolution that Hubble Space Telescope provides, we can see and learn from every star that makes up a pattern.”

More than just a stunning picture, each resplendent star reveals a wealth of valuable information and clues to the history of the galaxy’s formation.

Characteristics including star luminosity, color, and distribution provide insight into the dynamics, and evolution of a galaxy. For instance by analyzing star color, researchers can infer its age, and from luminosity its distance from Earth. The PHAT team used this information to develop star formation histories for M31, which are like fossil records in space, in that they provide clues to the physics that drove the evolution of the galaxy.

Models of stellar evolution allow the team to predict the specific ways stars of various ages, masses, and chemical compositions emit light. At the same time dust, small grains of interstellar matter, diminishes star light we see on Earth. Measuring the star formation history of M31 requires decoding the number of stars of each type (age, mass, chemistry) and how much dust is obscuring their light.

Applying this type of modeling to 100 million stars required powerful computation, so the team turned to XSEDE, a cyberinfrastructure that allows researchers to interactively share computing resources, data, and expertise. Through XSEDE, the team gained access to the Stampede supercomputer at TACC, one of the Top 10 most powerful systems in the world.

Stampede
[Learn More]

“We had to measure over 100 million objects with 100 different parameters for every single one of them,” Dalcanton said. “Having XSEDE resources has been absolutely fantastic because we were able to easily run the same process over and over again in parallel.”

"We received time on other supercomputers, but Stampede proved to be the most efficient for our needs. It turns out that the architecture works well with our codes -- they run faster than on other supercomputers."

-Daniel Weisz

By running software on Stampede, the team determined the ages of every star mapped, looked for patterns of star formation, and observed how the galaxy evolved over time.

“We received time on other supercomputers, but Stampede proved to be the most efficient for our needs. It turns out that the architecture works well with our codes — they run faster than on other supercomputers,” Weisz said.

The combination of the HST observations of Andromeda and modeling on Stampede allowed the PHAT team to solve a decades old problem about whether stars more massive than the sun form the same way everywhere in the universe.

Abi Saha

Abi Saha
National Optical Astronomy Observatory (NOAO)

Adam Leroy

Adam Leroy
National Radio Astronomy Observatory (NRAO)

Alexia Lewis

Alexia Lewis
University of Washington

Andy Dolphin

Andy Dolphin
Raytheon

Anil Seth

Anil Seth
Harvard‐Smithsonian Center for Astrophysics (CFA)

Antonela Monachesi

Antonela Monachesi
University of Michigan

Ata Sarajedini

Ata Sarajedini
University of Florida

Ben Williams

Ben Williams
University of Washington

Chris Kochanek

Chris Kochanek
Ohio State University

Claire Dorman

Claire Dorman
UC Santa Cruz

Cliff Johnson

Cliff Johnson
University of Washington

Daniel Weisz

Daniel Weisz
University of Minnesota

Dimitrios Gouliermis

Dimitrios Gouliermis
Max-Planck Institute for Astronomy (MPIA)

Dustin Lang

Dustin Lang
Princeton University

Eric Bell

Eric Bell
University of Michigan

Evan Skillman

Evan Skillman
University of Minnesota

Hans Walter-Rix

Hans Walter-Rix
Max-Planck Institute for Astronomy (MPIA)

Hui Dong

Hui Dong
National Optical Astronomy Observatory (NOAO)

Izaskun San Roman

Izaskun San Roman
University of Florida

Jacob Simones

Jacob Simones
University of Minnesota

Jason Kalirai

Jason Kalirai
The Space Telescope Science Institute (STScI)

Jason Melbourne

Jason Melbourne
CalTech

Jon Holtzman

Jon Holtzman
New Mexico State University (NMSU)

Julianne Dalcanton

Julianne Dalcanton
University of Washington

Karl Gordon

Karl Gordon
The Space Telescope Science Institute (STScI)

Karrie Gilbert

Karrie Gilbert
University of Washington

Keith Rosema

Keith Rosema
Random Walk Group

Kirsten Howley

Kirsten Howley
University of California, Santa Cruz (UCSC), Lawrence Livermore National Laboratory (LLNL)

Knut Olsen

Knut Olsen
National Optical Astronomy Observatory (NOAO)

Kris Stanek

Kris Stanek
Ohio State University

Leo Girardi

Leo Girardi
Osservatorio Astronomico di Padova

Lori Beerman

Lori Beerman
University of Washington

Luciana Bianchi

Luciana Bianchi
John Hopkins University

Martha Boyer

Martha Boyer
John Hopkins University

Morgan Fouesneau

Morgan Fouesneau
University of Washington

Paul Hodge

Paul Hodge
University of Washington

Phil Rosenfield

Phil Rosenfield
University of Washington

Rachel Wagner Kaiser

Rachel Wagner Kaiser
University of Florida

Raja Guhathakurta

Raja Guhathakurtar
UC Santa Cruz

Soeren Larsen

Soeren Larsen
Utrecht University

Stephanie Gogarten

Stephanie Gogarten
University of Washington

Tod Lauer

Tod Lauer
National Optical Astronomy Observatory (NOAO)

In their recent paper, the team provided the most conclusive evidence to date that shows star formation appears to be a universal process. That is, no matter how different the local conditions for star formation are, the resulting numbers and masses of stars appear to be the same everywhere in M31. Their finding has broad implications across all of astronomy — from how stars form to how astronomers interpret the light from distant galaxies.

With 100 billion galaxies in the Universe, there is much more work to be done to understand the infinite realm beyond our own. While the group has completed their Andromeda mapping project, they continue to chart other galaxies in the Universe.

“We do similar work on other nearby galaxies. Andromeda looks most like our own, but there are hundreds of nearby galaxies that look nothing like M31,” Wesiz said. “They’re interesting for exactly that reason, though. We’re trying to understand why galaxies have such diversity and what implications for the broader picture of how our universe has evolved.”


Credit: NASA, ESA, J. Dalcanton, B.F. Williams, and L.C. Johnson (University of Washington), the PHAT team, and R. Gendler. Content by Makeda Easter