Sometimes even the most powerful computer is no match for the human mind.

That is what a worldwide team of citizen scientists proved today, in results presented this week at the 237th meeting of the American Astronomical Society. More than 8000 citizen scientists from around the world joined together online to inspect galaxies from the Sloan Digital Sky Survey (SDSS) with Galaxy Zoo 3D, part of the Zooniverse citizen science platform.

"Galaxies are beautiful, complex objects, with structures like spiral arms that can vary from thick and solid to wispy and cloud-like," said Associate Professor of Physics and Astronomy Karen Masters, spokesperson for SDSS-IV and the lead author of this work. "If you ignore all this complexity and treat all galaxies as smooth blobs, you’ll miss important truths about how galaxies work."

This deeper look at galaxies is made possible by the spectra observed with the Mapping Nearby Galaxies at Apache Point Observatory (MaNGA) survey, one of the SDSS’s component surveys. Spectra are graphs that show the amount of light given off by a galaxy at different wavelengths of light, much like a rainbow shows the amount of sunlight in various colors. Most previous sky surveys have measured only one spectrum for an entire galaxy, but MaNGA works differently.

Rather than measuring only a single spectrum for an entire galaxy, MaNGA uses bundles of 19, 37, 61, 91 or 127 fibers spread right across the galaxy in a hexagonal shape. Thus, the MaNGA survey measures spectra at hundreds of separate points in a single galaxy, resulting in a “data cube” containing full spectroscopic information at each point in the galaxy. Researchers can then study each data cube to reveal its galaxy’s detailed chemical composition, find the ages and motions of the stars inside it, and map out the ionized gas within, which can in turn be used to estimate how many baby stars are being formed at each position in the galaxy.

The spectra MaNGA measure come from different structures in the galaxies. Images from the SDSS Legacy Survey imaging of all galaxies in the MaNGA target list were the input for Galaxy Zoo: 3D; from there, the volunteers took over.  Researchers used the crowdsourcing technique first popularized by the main Galaxy Zoo site, but instead of simply identifying features, volunteers were asked to draw over the image of the galaxy to show where specific features were.

"I'm continually impressed by the efforts of our citizen scientists, but even more so with the results of this more involved task," said Oxford University’s Chris Lintott, principal investigator of the overall Zooniverse project. "It suggests all sorts of new ways we could collaborate with them."

A full collection of around 15 drawings per galaxy, from 15 independent volunteers, were then combined into “heat maps,” which researchers from the team used to identify what visible features match with which spectroscopic features.

Galaxy Zoo: 3D’s citizen science technique allows the MaNGA data to be used to trace the impact of a galaxy’s internal structures on, for example, how stars begin to form in a galaxy and how the process of star formation comes to an end. One previously-published result from Galaxy Zoo: 3D was able to show how stars form in a galaxy’s spiral arm, and then fall behind as the arm passes through the galaxy. Another study showed how rings of star formation are common at the ends of galactic bars.

"Galaxy Zoo 3D has allowed us to do science that has never been possible before, by telling us just where to look for light from the different component parts that make up thousands of galaxies," said Michael Merrified, a professor at the University of Nottingham and a member of the MaNGA team who has made use of the Galaxy Zoo: 3D data.

One example of this is work presented by Haverford student, Shoaib Shamsi '20, who is also presenting a poster at the meeting. Shoaib’s work used the Galaxy Zoo: 3D information to measure the fraction of baby stars which are formed in a galaxy's spiral arms, compared to the galaxy as a whole. Finding it ranged from 40-60%.

It just goes to show that many eyes are better than a few, and computers won’t be replacing humans for everything any time soon.

Read about the other Haverford student, staff, and faculty presentations at the 2021 AAS meeting: https://www.haverford.edu/physics-and-astronomy/research/virtual-talks-presentations