Amidst the vast tapestry of the cosmos, galaxies reside in a variety of environments, each shaping their evolution in unique ways. Astronomers describe a galaxy's environment based on where it resides in the interconnected cosmic web, which can be thought of as the web a spider creates in a forest. Galaxy clusters gravitationally bind hundreds or even thousands of galaxies together. These clusters act as the junctions or nodes between the cosmic web and are akin to the intersections of a spider’s web. Filaments spanning millions of light years connect the nodes in the cosmic web, enabling the transport of gas and dark matter between them. Filaments can be thought of as the strands of the spider’s web. The holes in a spider’s web are the equivalent to void regions in the universe, where there is essentially nothing but a few galaxies that are sparsely spread out. The various components of the universe can be seen below in Figure 1. In the nearby universe, galaxies nestled in clusters, such as the Virgo cluster as seen in Figure 1, tend to exhibit significantly lower star formation compared to their counterparts adrift in the void regions. This divergence sparks an open yet tantalizing question: why do those galaxies evolve differently?

Supercomputer simulation image

Pictured at right is a supercomputer simulation of the large-scale structure of the universe. The yellow blobs depict various galaxy clusters or nodes, some of which are labeled. The black strands represent the cosmic filaments that connect clusters and transport material between nodes. The intervening gray blotches show voids where comparatively little matter exists. (Credit: Instituto de Astrofísica de Canarias)


To understand the large-scale structure of the universe better, astronomers need to study how nodes, or stable galaxy clusters, form from their precursors, dubbed protoclusters. As the infant stage of galaxy clusters, protoclusters have not yet had time to virialize, the process where orbits within the structure 'settle down' into a stable state. To date, only a few tens of protoclusters have ever been confirmed because they require many observations of individual galaxies within the protocluster to confirm association with the structure. Protoclusters first formed many billions of years ago, meaning they only exist in the distant universe, and thus also require powerful instrumentation to detect.

Roxana Popescu
Roxana Popescu

Roxana Popescu, a fifth year graduate student at UMass Amherst mentored by Professor Alexandra Pope, and her collaborators have recently studied a sample of about 200 candidate protoclusters, that they call the PC15 protocluster candidates. The PC15 protocluster candidates were first identified from infrared observations taken by the Planck Collaboration. Only candidates that appeared very bright in the infrared observations were selected. At these infrared wavelengths, objects that appear very bright can indicate that they contain copious amounts of dust and gas, the material necessary to make stars. Astronomers expect protoclusters to exhibit lots of active star formation since they contain relatively young galaxies, hence why the team selected infrared-bright sources.

From images taken by space telescopes like the Wide-field Infrared Survey Explorer (WISE) and Herschel, Roxana and collaborators measure the amount of light at different infrared wavelengths from each of the sources in the PC15 protocluster candidates. WISE traces the amount of mass, in stars, of the candidates, while Herschel traces the rate at which stars are forming, by essentially using the amount of dust present as a proxy. Both WISE and Herschel have imaging coverage of the protocluster candidates; individually faint galaxies within each image might not be detected, but when the images are averaged together, or stacked, the final result is a clear detection of the protocluster light. WISE may not be sensitive enough to detect individual protocluster galaxies, but stacking the images enabled the team to detect the total light from the protocluster candidates. Roxana recalls this process, “One moment that was really exciting was when we stacked the WISE observations and we actually saw a signal… we knew this sample would be detected [by Herschel]… but it was very nice to see that it was detected by WISE ”. From these averaged images of the PC15 protocluster candidates, this team measured the average total star formation rate and total mass in stars (the “stellar mass”) of the PC15 protocluster candidates.

By utilizing the stacking approach to obtain new measurements for the PC15 protocluster candidates, Roxana and collaborators could directly compare their findings with individually-detected dusty star-forming galaxies within other protoclusters. Through this analysis, they discovered notable differences in both the total rate at which stars form per year and the mass of stars in the galaxies. Specifically, they find that the PC15 protocluster candidates exhibit twice the star formation rate and four times the stellar mass compared to what is observed within individual galaxies in other protoclusters. This suggests that much of the star formation and stellar material of the protocluster candidates originates from protocluster member galaxies that are individually unseen. The team's stacking method has proven to be quite versatile and can be used to investigate the properties of other protocluster candidates to understand how much star formation occurs within protoclusters in the early universe. While more protocluster candidate observations are necessary to better understand these relationships, Roxana and collaborators' work may one day help reveal the childhoods of massive galaxy clusters like we see today in the nearby universe.

Open access to the full text of the original paper is available.

About the Author:

Moiz Khalil is a senior undergraduate studying astronomy at UMass Amherst. He currently studies the atmospheres and colors of brown dwarfs and wants to pursue a career in science communication. Outside of science, he enjoys listening to music and reading novels.