DESI launches five-year quest to understand the universe

Monday, May 17, 2021

The Dark Energy Spectroscopic Instrument (DESI), an international collaboration aiming to unravel the mystery of dark energy and fill in our 3D map of the universe, officially began on May 17.

There is a pattern printed on the fabric of spacetime.

Each piece of the pattern looks, in two dimensions, like a circle surrounded by a ring – as if some cosmic hand had thrown pebbles into the dense early universe, creating splash-points and ripples – then suddenly froze the pond.

The pattern isn’t easy to spot. The circles and rings are subtle, and they overlap in complex ways. But, crucially, they are all the same size. That’s because the “freezing of the pond” – the change in state when the early universe went from a plasma of charged particles to a neutral soup of mostly hydrogen – happened everywhere at once. The light known as the cosmic microwave background was released in that instant, and without the light to scramble it, the circle-and-ring pattern of mass was frozen in place.

Baryon acoustic oscillation

The baryon acoustic oscillation is a large-scale pattern of mass. You’re slightly more likely to find galaxies on the circles and rings rather than elsewhere. This artist’s conception simplifies and exaggerates the subtle pattern. (Credit: Gabriela Secara, Perimeter Institute)

 

Since then, the circle-and-ring pattern has stretched out because spacetime itself is stretching – the way a pattern printed on a balloon will stretch as the balloon inflates.

This pattern is called the Baryon Acoustic Oscillation. A new project is setting out to map it – and in so doing, creating a powerful new tool for understanding the expansion of spacetime, and the mysterious dark energy that drives it.

Introducing DESI

The new project is called DESI, the Dark Energy Spectroscopic Instrument – an international collaboration managed by the US Department of Energy’s Lawrence Berkeley National Laboratory. DESI’s five-year quest to understand the history and evolution of spacetime officially got underway May 17 at Kitt Peak National Observatory, near Tucson, Arizona, in the United States.

“DESI is a new scientific instrument, on a retrofitted telescope, that measures the distances to galaxies,” says Perimeter computational scientist Dustin Lang, who is one of the researchers who did the intense software work and data collection required to make DESI’s observations possible. There have been surveys of galaxies before – most recently the Sloan Digital Sky Survey – but DESI is more powerful than Sloan by a factor of 20.

“All the elements of DESI’s design are pushed toward doing a lot of measurements, fast,” says Lang. “We mass-produce galactic data.”

The instrument captures light from the telescope like a digital camera – but with two crucial differences. First, the camera is only interested in a few of the pixels in the image plane: the ones exactly corresponding to distant galaxies and quasars. To isolate them, tiny motors, guided by sophisticated software, position 5,000 fibre optic cables so that each one is aligned with a single galaxy.

Second, the instrument doesn’t simply capture light, it breaks light down like a prism breaking down a sunbeam, into a rainbow called a spectrum. “On a good night, we can get 5,000 spectra every 20 minutes,” says Lang. “Previous sky surveys collected about four million spectra in total – and we can get more than 100,000 a night.”

Adding depth to a billion-galaxy map

DESI’s official launch follows a truly staggering piece of prep work: the largest ever map of the sky, combining 1,400 nights of observation from three telescopes, and containing two billion stars and galaxies.

The map, called the Legacy Sky Survey, debuted at the January 2021 meeting of the American Astronomical Society, but if you missed it, you can browse it online. Allow some time: the massive map contains trillions of pixels.

It might seem reasonable to ask: if that’s the prep work, what’s left for DESI to do? The answer: the Legacy Sky Survey maps the angular position of its billion galaxies – that is, where they are on the sky in two dimensions. It’s this map that allows researchers to precisely point all 5,000 of those fibre optics. But as DESI surveys a subset of these galaxies in more detail, it will literally add a new dimension to the map: depth. It will measure how far away each targeted galaxy is.

Specifically, the DESI researchers will take a detailed spectral measurement of each galaxy. That data will let them assign each galaxy a redshift.

Redshift is a kind of doppler effect for light: when an object is moving toward you, the waves bunch up, creating higher frequency, bluer light. When an object is moving away from you, the waves stretch out, shifting the spectrum toward the red. Since the 1920s, it’s been known that the farther away a galaxy is, the faster it is receding. Distant galaxies are very red-shifted, nearer ones less so.

It’s also well understood that distant galaxies are distant not just in space, but in time. If the light from a distant galaxy takes, say, five billion years to reach Earth, then what we see is that galaxy as it appeared five billion years ago. By studying very red-shifted galaxies, we study the early universe.

In other words: DESI is a time machine.

Using spectra to slice the sky

DESI will collect data for five years. “We aim to observe on the order of 30 million galaxies,” says Will Percival, Distinguished Research Chair in Astrophysics at the University of Waterloo, where he is the director of the Waterloo Centre for Astrophysics and cross-listed with Perimeter Institute.

Thirty million is only a subset of the one billion galaxies in the Legacy Sky Survey map, but it will be a tenfold improvement on the Sloan Digital Sky Survey, which has collected about 3 million galaxy spectra in over 10 years of observations.

Also targeted are about 2.5 million quasars, which are bright objects found at the centre of some galaxies, powered by gas spiraling at high velocity into an extremely large black hole. Quasars are so bright that they outshine all the stars in their host galaxies, and they’re important in this context because they are the most distant objects DESI can image. Studying quasars will let DESI take a snapshot of the universe as it existed 11 billion years ago.

The big dataset of galaxies and quasars, each with redshifts assigned, will let researchers slice the sky into layers. Each layer represents a different moment in the universe’s history.

Baryon Acoustic Oscillation

The baryon acoustic oscillation has a characteristic size, which changes only because spacetime itself is stretching. By studying the pattern at different points in the universe’s history, researchers chart the expansion of spacetime. The expansion is speeding up, driven by the mysterious force known as dark energy. (Credit: Gabriela Secara, Perimeter Institute)

 

Circling back to the circles-and-rings

Once DESI has sliced the sky into layers of time, they can begin the tricky process of looking for the circle-and-ring pattern of the Baryon Acoustic Oscillation.

That’s the moment Percival and his team start looking to extract the cosmological signal from the galactic clustering. The signal will be subtle – looking out from one galaxy, you’re less than one percent more likely to find another galaxy on the ring pattern around it, rather than at slightly smaller or larger separations – so the big dataset is necessary. But on the bright side, Percival says, the technique they use is robust: “That means it’s insensitive to systematic errors – so that it’s very difficult to do the experiment wrongly.”

Researchers will look for the circle-and-ring pattern appearing at various points in the universe’s evolution, from 11 billion years ago to nearly the present day. They predict that they will see the pattern stretch out as spacetime itself stretches out: again, think of a pattern printed on a balloon.

Their ultimate goal is to study the history of that stretching. We know that spacetime stretches outward from the explosion of the big bang. We know that is it is pulled inward by its own gravity. Based on those two factors alone, one would expect the universe to either expand at a fairly constant rate, or for the expansion to be gradually slowing. But in 1998, scientists discovered that the expansion of spacetime is actually speeding up. Something extra is spreading out spacetime itself. Scientists named that extra something “dark energy.”

“This dark energy question is one of the most exciting questions in physics at the moment,” says Percival. “An analogy would be seeing the effects of electricity – the sparks and shocks – but not knowing the physics required to explain them. Without the physics, it would just seem like magic.”

DESI scientists will study how the circle-and-ring pattern stretches over time. It’s a new window into the history of dark energy.

The possibility of discovery doesn’t end there. It’s perhaps worth remembering that dark energy itself was discovered in a survey of distant galaxies. No one was looking for it – it emerged from the dataset as a complete surprise.

“Any time you undertake an experiment that is 10 or 20 times more powerful than you had before, you open up a huge amount of potential for new discoveries,” says Percival. “We might find something unknown, something that doesn’t follow our model of cosmology. Or a new kind of object, a new feature, something we haven’t seen before.

“We’re scientists – it’s the unknown that excites us.”

The DESI collaboration is honoured to be permitted to conduct scientific research on Iolkam Du’ag (Kitt Peak), a mountain with particular significance to the Tohono O’odham Nation.

  1. 2021 (6)
    1. June (1)
    2. May (1)
    3. April (1)
    4. March (2)
    5. January (1)
  2. 2020 (14)
    1. December (1)
    2. October (1)
    3. September (1)
    4. August (2)
    5. July (1)
    6. May (2)
    7. April (3)
    8. March (2)
    9. January (1)
  3. 2019 (12)
    1. October (1)
    2. September (1)
    3. July (4)
    4. June (5)
    5. April (1)
  4. 2018 (2)
  5. 2017 (5)
  6. 2016 (3)
  7. 2015 (3)
  8. 2014 (2)