Far-Infrared Data Validates Galaxy Evolution Models
UMass Amherst post-baccalaureate finds that galaxy properties from visible-light data remain largely robust even when hidden dust emission is included.
While astronomers have mapped much of the universe through visible-light observations, some of its most actively star-forming galaxies remain hidden behind thick clouds of cosmic dust. To peer into these obscured regions, researchers must look beyond the visible spectrum and into the far-infrared, where dust-enshrouded star formation can be detected directly. By combining data from space telescopes spanning the electromagnetic spectrum, UMass Amherst post-baccalaureate researcher Seamus McNulty investigated how galaxies formed stars across billions of years of cosmic history. By adding far-infrared data, which reveals an otherwise hidden, dust-obscured component of galaxies, McNulty showed that star-formation measurements derived from visible-light observations hold up remarkably well."
McNulty joined Professor Whitaker's group at UMass Amherst in 2022 as an undergraduate, inheriting a project from a departing postdoctoral researcher. The work grew into his Honors Thesis, which he completed before graduating in 2024. He has continued this research as a post-baccalaureate, while working full time in the tech industry.
The project McNulty inherited tackled a long-standing measurement problem. Visible-light surveys have catalogued star formation in thousands of galaxies, but in the dustiest systems -- often the most prolific star-formers -- that light never escapes. Quantifying what those surveys miss requires far-infrared data, where dust grains heated by young stars re-radiate the absorbed light."
To tackle this problem, McNulty drew on several major astronomical surveys that imaged galaxies across different wavelengths of light. Existing catalogs from the 3D-HST and CANDELS surveys, which together provide deep spectroscopy and imaging of distant galaxies. These catalogs already contained ultraviolet through mid-infrared observations for tens of thousands of galaxies, but lacked direct measurements in the far-infrared where dust emission is strongest. To extend these catalogs into the far-infrared, McNulty added observations from the Herschel Space Observatory, allowing him to study galaxies across 70 - 350 microns.
One of the most significant challenges in this work came from the lower resolution of the Herschel images. In crowded images, light from multiple sources can blend, making it difficult to determine which galaxy is actually emitting it. For reference Herschel's resolution is roughly 100 times coarser than Hubble's, so a single far-infrared blob often corresponds to several optical galaxies. Before analyzing the galaxies themselves, McNulty had to painstakingly comb through the catalog and cross-reference objects between surveys to correctly attribute the blended far-infrared light to individual galaxies.
Once the catalog – which the team called 3D-Herschel – was ready, UMass graduate student Aubrey Medrano used Bayesian modeling techniques to estimate physical properties such as stellar mass, dust content, and star formation rates for over 40,000 galaxies spanning the past ten billions of years of cosmic history. By comparing fits that included Herschel’s far-infrared photometry against fits limited to shorter wavelengths , McNulty could test exactly which galaxy properties astronomers can recover without far-infrared data and which they cannot. The headline result was reassuring: stellar masses, ages, and star formation rates derived without the far-infrared data agreed with previous measurements within typical uncertainties, and the galaxies' star-forming main sequence – a key benchmark in galaxy evolution – shifted by less than 0.1 dex.
The results provide new confidence in existing models of galaxy evolution while also highlighting important cases where far-infrared observations can reveal details that shorter wavelengths alone may miss. The dust itself, however, tells a different story. Without far-infrared data, dust temperatures were systematically inferred to be about 7 K colder at all epochs, and the relationship between mid- and total-infrared brightness varied with stellar mass in ways shorter-wavelength data couldn't capture. The takeaway: optical and mid-infrared observations are enough to characterize a galaxy's stars, but understanding its dust, and the obscured star formation buried within it, still requires far-infrared coverage. Realizing that the future will require new instruments, UMass Amherst astronomer Alexandra Pope, a member of McNulty's thesis committee, is helping make the case for one — PRIMA, a NASA Probe-class mission concept.
Reflecting on the project, McNulty described the experience as both intimidating and deeply rewarding. Taking over such a comprehensive study meant diving into astronomical catalogs and teaching himself the technical skills required to carry the work forward. The scale of the project was daunting at first, but overcoming those challenges became one of the most fulfilling parts of the experience. His advisor watched that growth firsthand. “Seamus is one of the best students I've ever worked with,” Professor Whitaker remarked. “Over four years, he inherited a complicated project, taught himself the analysis, and grew into a researcher who could make technical calls I would normally have to make myself. Watching that kind of independence develop has been one of the real joys of advising.” What makes the work especially meaningful to McNulty is the idea that the catalog will continue to support future research.. With the paper set for resubmission this month and the catalog soon to be released alongside it, McNulty sees many open questions about galaxy evolution and dust-obscured star formation still to explore. One area he is particularly interested in exploring further is how assumptions about dust temperature influence galaxy models and inferred physical properties.
Throughout the post-baccalaureate position, McNulty balanced research alongside a full-time position in the tech industry, giving him a rare perspective on both industry and academia. While he appreciated the experience of working in both environments, he found the collaborative and discovery-driven nature of astronomy especially fulfilling and hopes to continue pursuing research in the future. Looking back, his advice to younger students entering the field is simple: learn how you work best, stay organized, and plan ahead without becoming overwhelmed by the future. “Take things slow,” he said. “Have a plan, but don’t look too far ahead or you’ll miss the scenery along the way.”
About the author: Dan Kidwell is a senior undergraduate studying astronomy and physics at UMass Amherst. He currently works in high-energy astronomy, conducting simulated observations of stellar winds in the Galactic Center. Outside of his studies, he enjoys hiking, camping, and photographing the night sky.