Modeling the Cosmic Web

Modeling the Cosmic Web

When you look up at the sky at night, you probably see a smattering of stars – perhaps the Big Dipper and other bright ones. If you live in a place with very little light pollution, you may even see the Milky Way streaking across the sky.

But if you were able to see the universe as astrophysicists and computer scientists see it, you would see a vast cosmic web stretching into the distance. There would be galaxies and dark matter dotting the sky like intersections in a spiderweb, connected by strands of cosmic gasses, and voids of nothingness in-between. That’s what astrophysicists see with the help of powerful telescopes and supercomputers that analyze the data from them. Computer scientists at the Department of Energy’s Lawrence Berkeley National Laboratory (Berkeley Lab) are developing software to enable that analysis.

Astrophysicists know a great deal about the universe. But there are two major, outstanding mysteries: dark matter and dark energy. Visible matter, including galaxies, makes up approximately 5 percent of the universe and includes all matter that gives off light and that we can detect with modern instrumentation. Dark matter makes up about 25 percent of the universe. Dark matter is a form of unseen matter that we are not able to directly detect or measure. It appears to only interact with visible matter via gravity. We only know it exists because of the strange way galaxies and stars move. Dark energy makes up the remaining massive 70 percent of the universe. We know even less about dark energy than dark matter. It’s the name that astrophysicists give the force that is making the universe not only expand but expand ever faster. One of the few things we have observed about dark energy is that it counterbalances gravity, which pulls the universe together.

The Dark Energy Spectroscopic Instrument (DESI) is giving us new insights into these cosmic phenomena. DESI started collecting survey data in 2021, at the Kitt Peak Observatory in Arizona. It will collect data until 2026, during which time it will observe an estimated 40 million galaxies and quasars. Quasars are incredibly bright cosmic objects. As the light from quasars travels through the universe, the hydrogen gas that exists in space absorbs it at specific wavelengths. By studying this light, scientists can understand the patterns of hydrogen and better pinpoint the pattern of dark matter through the effects of gravity. Mapping dark matter helps scientists narrow down the possible forms it could take, like a detective eliminating suspects. Scientists estimate that DESI will collect data on nearly three million quasars, including over 800,000 quasars that will be sufficiently distant to probe the hydrogen content of the universe. That’s more than five times as many distant quasars as astrophysicists have been able to collect data on before.

A team based at DOE’s Berkeley Lab is building software that can analyze and model hundreds of thousands of light patterns from these quasars. One of the main pieces of software they’ve developed is called Nyx – appropriately named for the Greek goddess of the night. One function of the software is to predict the anticipated observational data given current theory. If the observational data show something different, it suggests either there is a problem with the observations, or there’s a gap in theory. Scientists run this software on incredibly powerful supercomputers supported by DOE. In the past, the lab has conducted this analysis on the computers at the National Energy Research Scientific Computing Center (NERSC), an Office of Science user facility run by Berkeley Lab. The team is now extending the software to take full advantage of our most powerful exascale supercomputing facilities.

To understand the vastness of our universe, tools that take observations aren’t enough. We also need the powerful computational tools that astrophysicists can use to make sense of massive amounts of data.

Originally published at

The post Modeling the Cosmic Web first appeared on Social Gov.

This story originally appeared at Tech - Social Gov