UW Researcher Contributes to Sky Surveys in Preparation for Dark Energy Spectroscopic Instrument

When people go camping in Wyoming, they often look to the sky to see the moon and the stars and, perhaps, ponder their place in the universe.

For the past five years, Adam Myers, a University of Wyoming associate professor of physics and astronomy, has been part of a worldwide research team that worked painstakingly to gather 2-D images of more than 1 billion galaxies -- images that may well provide some of those answers.

It took three sky surveys -- conducted at telescopes on two continents, covering one-third of the visible sky and requiring almost 1,000 observing nights -- to prepare for a new project that will create the largest 3-D map of the universe’s galaxies and glean new insights about the universe’s accelerating expansion over the last 10 billion to 12 billion years.

The Dark Energy Spectroscopic Instrument (DESI) project, now finishing construction in Arizona, will explore this expansion, driven by a mysterious property known as dark energy, in great detail. It also could make unexpected discoveries during its five-year mission. DESI is expected to be online in spring 2020, with the project running through 2025.

A collection of 5,000 swiveling robots, each carrying a fiber-optic cable, will point at sets of preselected sky objects to gather their light (see a related video) so it can be split into different colors and analyzed using a series of devices called spectrographs.

“Light travels 11 million miles a minute. DESI will make a map 10 billion times larger than light travels in a year,” Myers says. “These are the largest maps that humankind can make and, after the DESI survey, they will be in 3-D. We will know the distances to objects as well as the distance across the sky.”

Scientists will select about 33 million galaxies and 2.4 million quasars from the larger set of objects imaged in the three surveys. Quasars are the brightest long-lived objects in the universe and are believed to contain supermassive black holes. DESI will target these selected objects for follow-up spectroscopy after its launch.

A supercomputer, dubbed Cori, located at the National Energy Research Scientific Computing Center (NERSC) at Lawrence Berkeley National Laboratory, will be used to process the millions of spectra observed by the DESI instrument. This processing will extract the 3-D information, or distances, from the spectra, which cover a large range of the spectrum of light, Myers says.

Measurements of the spectra of galaxies from DESI will not only provide details about their distance of galaxies from Earth, but also about how fast they are accelerating through our universe.

“This will tell us something, potentially, about how gravity works and determine whether Einstein’s theory of relativity needs to be extended,” says Myers, who has served as the DESI targeting manager since 2016 and is funded by the Department of Energy.

Targeting involves determining which of the billions of the galaxies in the 2-D images of the sky will receive DESI follow-up spectroscopy to measure their 3-D distances. Myers has written tens of thousands of lines of software in pursuit of this goal.

“To turn a raw image of the sky into something useful requires all kinds of code,” Myers says. “We took different photos of the sky to try to align them. We want a uniform sample of galaxies across the universe -- some close, some far.”

Much like assembling a puzzle, Myers says individual images of the sky and objects have to be looked at closely to match their properties. For example, some images may have slightly more moonlight. Others may have been taken in slightly better, or worse, weather conditions.

“When people traditionally use a supercomputer, they use it to model stuff. We care about data,” Myers says. “We had to work with them (NERSC) to modify their systems to work with our data. We’re using a supercomputer in a new way.”

Collecting the Images

The initial, raw images were captured at three observatories, one located in Chile and two in Arizona.

“My group conducted a significant fraction of the imaging observations, despite being one of the smaller institutions in the collaboration,” Myers says of observations he made at the Mayall telescope at Kitt Peak National Observatory in Tucson, Ariz., and Cerro Tololo Inter-American Observatory in northern Chile. “My research group observed for a total of 44 nights.”

The surveys, which wrapped up in March and involved more than 100 researchers, have amassed images of more than 1 billion galaxies and are essential in selecting celestial objects to target with DESI.

New software designed for the DESI surveys, and precise positioning equipment on the telescopes, has helped to automate the image-taking process, setting the exposure time and filters, and compensating for atmospheric distortions and other factors that can affect the imaging quality.

The data from the surveys was routed to supercomputers at NERSC, which will be the major storehouse for DESI data. Supercomputers Cori and Edison were used to process the images of the sky from telescopes and extract the few billions of sources from them.

“Just processing the data required us to use millions of hours on one of the most powerful supercomputers in the world,” Myers says.

The latest batch of imaging data from these surveys, known as DR8, was recently publicly released, and an online Sky Viewer tool provides a virtual tour of this data. A final data release is planned later this year.