One Image, One Million Galaxies

One of the Hubble Space Telescope’s most iconic images is the Hubble Ultra Deep Field, which unveiled myriad galaxies across the universe, stretching back to within a few hundred million years of the Big Bang. Hubble peered at a single patch of seemingly empty sky for hundreds of hours beginning in September 2003, and astronomers first unveiled this galaxy tapestry in 2004, with more observations in subsequent years.

NASA’s upcoming Nancy Grace Roman Space Telescope will be able to photograph an area of the sky at least 100 times larger than Hubble with the same crisp sharpness. Among the many observations that will be enabled by this wide view on the cosmos, astronomers are considering the possibility and scientific potential of a Roman Space Telescope “ultra-deep field.” Such an observation could reveal new insights into subjects ranging from star formation during the universe’s youth to the way galaxies cluster together in space.

Roman will enable new science in all areas of astrophysics, from the solar system to the edge of the observable universe. Much of Roman’s observing time will be dedicated to surveys over wide swaths of the sky. However, some observing time will also be available for the general astronomical community to request other projects. A Roman ultra deep field could greatly benefit the scientific community, say astronomers.

As an example, a Roman ultra-deep field could be similar to the Hubble Ultra Deep Field – looking in a single direction for a few hundred hours to build up an extremely detailed image of very faint, distant objects. Yet while Hubble snagged thousands of galaxies this way, Roman would collect millions. As a result, it would enable new science and vastly improve our understanding of the universe.

One of the greatest challenges of the Roman mission will be learning how to analyze the abundance of scientific information in the public datasets that it will produce. In a sense, Roman will create new opportunities not only in terms of sky coverage but also in data mining.

A Roman ultra-deep field would contain information on millions of galaxies – far too many to be studied by researchers one at a time. Machine learning — a form of artificial intelligence — will be needed to process the massive database. While this is a challenge, it also offers an opportunity.