The Lynx Arc: A starburst 12 billion years ago

Want to take a peek at 25,000 of the most violent stars in the universe? Look no further than the Lynx Arc.

The most important scientific discoveries often arise from the humblest beginnings. For instance, the cosmic microwave background, a key piece of evidence for the Big Bang, was originally observed as an unexplained piece of radio noise. Similarly, the first known pulsar appeared as a modest periodic blip in Jocelyn Bell’s data. Even the faintest signal can spark a new field of inquiry.

An important — though perhaps less well-known — and similarly serendipitous discovery occurred in 2001, sparked by a faint curve in an image from the Hubble Space Telescope. A group of astronomers were observing a group of galaxies called RX J0848+4456, consisting of a cluster emitting x-rays and a less active cluster closer to Earth. Hubble observations showed several arcs around the cluster, possible indications of gravitationally lensed sources far beyond.

One of those structures, today referred to as the Lynx Arc, is thought to be a treasure trove of hot, massive stars, formed 12 billion years ago and seen now as they were in their infancy. While discerning individual stars may be beyond our current capabilities, studying the Arc itself can tell us many things about star formation in the relatively early universe, and provide fascinating insights into populations of exotic stars.

Real clusters have curves

Galaxy clusters are great targets for studies of gravitational lensing. Large galaxy groups, especially those with large amounts of dark matter, are massive structures that can bend light quite dramatically. In fact, measurements of lensing are a key tool when it comes to weighing these clusters. Although the precise shape of a lensed object depends on the target cluster, most distant galaxies focused by gravitational lenses appear as arcs or fine curves.

When observing RX J0848+4456 using Hubble and the Chandra X-ray Observatory, Holden et al. (2001) noticed a number of arcs surrounding the group of galaxies, most likely lensed by the closer, less active cluster. They were able to confirm one of the most dramatic arcs as a gravitationally lensed object far beyond the group.

The measurements weren’t fine enough to glean many details about the mysterious object, but the presence and absence of specific emission lines seemed to rule out particular candidates. Strong [O III] lines implied that gas was being ionized by a low-metallicity black body of temperature 80,000–100,000 K, and weak He II and N V lines seemed to rule out an active galactic nucleus. Regardless of the identity of the arc, the authors were optimistic about its use as a diagnostic for the mass of the closer cluster.

Further Hubble observations by Fosbury et al. (2003) appeared to confirm some of the first group’s suspicions. They found that the emission spectrum was unlike any previously observed, and appeared to indicate strong photoionization of gas by a hot source — likely a star cluster. The group was able to model the cluster photoionization as coming from a single hot object with a temperature of 80,000 K, and found that it fit the data fairly well.

The emission lines and high temperature were indicative of a group of hot stars, likely formed in the early universe. The observed metallicity of the gas was low, about 5% of the solar metallicity, and this lack of heavy elements fit well with the source’s high redshift. If it formed billions of years ago, the gas could not have been enriched much by supernovae and earlier generations of stars — at least, compared to the present day.

Modeling the Hβ emission gave a rough estimate of the rate of production of ionizing photons, and stellar population modeling implied that the source consisted of about 1 million massive stars, an extraordinary quantity. All of these signs seemed to point towards a cluster of newly-formed Population III stars, objects that have been theorized for decades but have yet to be definitively detected. The sheer size of the cluster was quite impressive, as very few massive stars are known to exist.

Are Wolf-Rayet stars responsible?

One problem that arose from the measurements by Fosbury et al. was that normal O stars could not explain was He2+ emission; population models dramatically under-predicted the observed line strengths. Additionally, the assumed temperature (~80,000 K) was much larger than would be expected for typical O stars. A solution was proposed by Villar-Martin et al. (2004): Wolf-Rayet stars.

Wolf-Rayet stars are objects with little or no hydrogen at their surfaces but strong lines corresponding to other elements, such as nitrogen, oxygen, carbon and helium. They are some of the hottest stars in the universe, and exhibit strong stellar winds and dramatic mass-loss rates. Bright and massive, they are thought to be common supernova progenitors. The spectra from a collection of Wolf-Rayet stars would be sufficient to ionize He+ to He2+, the group concluded. A set of about 25,000 stars would cause the necessary photoionization.

One problem remained: a dearth of continuum emission. Sources that produced the required Hβ line luminosity predicted a continuum 10 times as strong as the upper limits of the observed continuum. Villar-Martin et al. considered several solutions to the problem:

• An active galactic nucleus, which would require broader lines and weaker [Si III] emission than were observed.
• A “hidden starburst”, where continuum emission would be blocked by large quantities of dust.
• A difference in the spatial distribution of the stars and the ionized gas, causing the nebula emission to be amplified by lensing much more than the continuum emission.
• A set of Population III stars, as considered by Fosbury et al., which would not fit the nebular abundances of heavy elements.

The group considered the third explanation to be the most viable, ruling out the others by virtue of abundance and line strength discrepancies. Overall, their Wolf-Rayet star hypothesis appeared to successfully solve the ionization problem.

The rarest stars

The existence of such a large population of Wolf-Rayet stars would be remarkable. Only a few thousand are expected to exist in the Milky Way, and most galaxies — perhaps with the exception of Wolf-Rayet galaxies undergoing periods of starbursts — should also have similarly low numbers. Even Fosbury et al.’s proposal of 1 million O stars would be significantly rare.

Starburst galaxies — for instance, Messier 82 — have been discovered closer to the Milky Way, and are active targets of study. However, high-redshift star-forming regions like the Lynx Arc provide us with information about stellar populations in the early universe. Even if these stars aren’t Population III stars, they aren’t far removed, and studying the abundances can hopefully tell us about the supernovae that enriched the nebula they’re embedded in, and therefore the first stars that formed in our universe.