Skip to main content

The Swift GRB Host Galaxy Legacy Survey: Understanding the Link between GRBs and Star Formation from the Nearby to the Early Universe

Periodic Reporting for period 1 - SHOALS (The Swift GRB Host Galaxy Legacy Survey: Understanding the Link between GRBs and Star Formation from the Nearby to the Early Universe)

Reporting period: 2015-05-15 to 2017-05-14

The Universe has changed tremendously over thirteen billion years of cosmic history. Shortly after the Big Bang the cosmos was a nearly-featureless ocean of hydrogen and helium: the development of structure and heavy elements, prerequisites for the formation and evolution of planets and life, required much longer time scales. To understand our origins we would like to observe how this process unfolded across time. Astronomers do this by exploiting the vast size of the universe and the finite travel time of light: by observing galaxies billions of light-years away we see how the universe appeared billions of years ago. But this is difficult: galaxies at these vast distances appear very dim, making them exceedingly difficult to study in detail.

To get around this difficulty, the SHOALS project relies on a tracer much more luminous than a typical galaxy whose intrinsic power can compensate for the effects of cosmic dimming. Long-duration gamma-ray bursts (GRBs) are produced by the explosions of rare, ultra-massive stars inside galaxies, and for a few minutes these events produce brilliant light at all electromagnetic wavelengths which can be studied in detail to reveal the presence, distance, chemical composition, and other properties of the galaxy in which it occurred. Even galaxies that would otherwise be too faint to be detected by the most powerful existing telescopes can be revealed by this technique. In this way, I and my collaborators pin-pointed a population of 119 galaxies which hosted GRBs during the past decade. We used information provided by observations of the burst while it was bright, as well as observations of the galaxy alone after the burst had faded, acquired via a multi-year, multi-facility observational campaign.

Our immediate objective has been to build a catalog of the properties of GRB-sampled galaxies spanning the Universe's history, and use this catalog to accomplish several scientific objectives. First, we sought to better understand the composition of the universe at each epoch (such as the numbers of small, faint galaxies versus the number of large, fully-formed galaxies) and infer how rapidly this changed over the Universe's history. Second, we have attempted to better understand how well galaxies selected by our GRB technique match the galaxies selected by traditional astronomical surveys, in order to assess the advantages and disadvantages of each method. In particular, we have aimed to understand whether GRBs happen specifically in certain types of galaxies while avoiding others, perhaps because the production of the types of stars that produce these bursts requires particular conditions (such as a low concentration of heavy elements).
"The lynchpin of the survey is the Spitzer Space Telescope, which observes at mid-infrared wavelengths and is sensitive to the light from ""typical"" (older) stars and therefore can be used to measure a galaxy's total mass in stars. Using data from this telescope we were able to measure the stellar masses of all galaxies within the SHOALS sample and study how the mass distribution of GRB hosts evolves with time. We demonstrated that this distribution evolves relatively little from a redshift of z=5 (12.5 billion years ago) to z=1.5 (9 billion years ago) and that throughout this period the majority of galaxies are relatively luminous. However, between z=1.5 and z=0 (9 billion years to the present) the mass distribution changes rapidly: the largest and most massive galaxies disappear. We attribute this to sensitivity of the GRB progenitor to the presence of heavy elements (""metals""), which build up in galaxies over time. Specifically, we demonstrate that if gamma-ray bursts cannot occur in galaxies with a mean metal abundance higher than that of the Sun, both the mass distribution (and its time evolution) and the redshift distribution of our sample can be fully explained. These results were published in a pair of papers: Perley et al. 2016a (ApJ 817:7) describes the redshift distribution, and Perley et al. 2016b (ApJ 817:8) describes the mass distribution.

In addition to studying stellar masses with Spitzer, we also seek to analyse other properties of the galaxies: for example, how rapidly they are forming stars, or how much interstellar dust they contain. This will enable us to study how these properties change with cosmic time, and will also help us test whether GRBs might occur more readily in some galaxies over others for reasons not related to metal content. This requires additional observations at different wavelengths other than those probed by Spitzer, and while a great deal of data had already been compiled at the start of the project, the the data set was not complete at that time. New observations carried out over the course of the project at the Keck, VLT, and Magellan optical telescopes have now completed this data set and analysis of these data is underway.

We also acquired observations of smaller, targeted subsets of the sample at millimeter wavelengths, radio wavelengths, and using near-infrared spectroscopy. So far we have completed two studies with these data: in Perley et al. 2017a (MNRAS 465:970) we show that most gamma-ray burst host galaxies have moderate total star-formation rates, while in Perley et al. 2017b (MNRAS 465:89) we show that a report in pre-existing literature of a GRB host galaxy with above-solar metallicity (which was in tension with our Spitzer result, above) was actually a misidentification with a foreground galaxy.

Beyond the immediate objectives of the survey, we also have been pursuing investigations of GRB hosts from other observational campaigns (beyond SHOALS) and other types of luminous explosion which might have cosmic utility (such as luminous supernovae). The PI has also remained active in identifying and studying new gamma-ray bursts that exploded during the project, joining the pan-European X-shooter GRB collaboration. The published works most directly relevant to the SHOALS core science case are given in the Publications list. In total, the PI has co-authored 39 refereed publications which appeared or were first submitted during the project, and many additional works are in preparation or have been submitted for publication."
Our GRB host galaxy survey is the largest effort of its type by far. The large size and unbiased selection methodology of the survey have allowed it to be the first to be able to directly and quantitatively address changes in the galaxy population over cosmic history using GRBs. Our study provides strong support for the consensus model of galaxy evolution in the Universe, under which large galaxies were able to assemble rapidly within a few billion years of the Big Bang, after which continued star-formation and metal-production continued at a gradually decreasing level from then until the present epoch. The conclusion of a solar-metallicity ceiling on GRB production is a significantly higher threshold than some other recent, smaller studies had suggested and provides support for the continued use of GRBs as galaxy probes in the future. It also challenges our theoretical understanding for how to produce rapidly-rotating stripped stars of the type that are thought to be needed to explode as GRBs, since basic theories had previously predicted that this threshold would be much lower.
Mosaic of Spitzer 3.6 micron observations of the 119 galaxies of the primary SHOALS sample.