Thursday, November 21, 2013

Revealing galaxies beyond our sight

The secrets of the universe can be revealed in many different ways. Serendipity often plays a role, where astronomers use unexpected events to make discoveries. Here's a striking example. In a few cases, astronomers have been able to map the positions and distances of galaxies that are so remote and faint they cannot be seen using our most powerful telescopes. However, the spectacular explosion of a single star can give away the galaxy's presence.

For context, I’ll explain what the limits of our telescopic vision are for distant galaxies. A result published in Nature (arXiv paper) a few weeks ago by Steven Finkelstein and collaborators claimed the title of the most distant galaxy with a securely known distance (other possibilities for more distant galaxies have been claimed, but with less secure distance estimates). The galaxy is called “z8_GND_5296”, which is a candidate for the most unappealing galaxy name. The light seen from z8_GND_5296 was emitted when the universe was only about 700 million years old, or about 5% of its current age. The galaxy was also found to be forming stars at a prodigious rate.

A Hubble Space Telescope image of z8_GND_5296, which is claimed to be the most distant galaxy with a securely known distance. Credit: V. Tilvi, S.L. Finkelstein, C. Papovich, A. Koekemoer, CANDELS, and STScI/NASA
This is an impressive result, but it doesn't represent the record for the most distant object with a securely known distance. That record is likely held by an exploding star that was seen when the universe was only about 630 million years old, as published in Nature papers in 2009 by Nial Tanvir et al (arXiv paper) and Ruben Salvaterra et al (arXiv paper). Here’s a press release describing the discovery.

The center of this image shows the afterglow of an exploded star which is claimed to be the most distant known object in the universe. The image was obtained from the Gemini-South and the Very Large Telescope. Credit: A. J. Levan.
A deep Hubble Space Telescope (HST) search for the galaxy containing this exploded star came up empty, a testimony to the incredible light show that massive stars can perform when they explode. The burst of gamma rays and other radiation from these "gamma-ray bursts" (GRBs) can provide our only clue about the very distant galaxies they occurred in. Once this radiation fades, any information about the galaxy fades with it. To turn things around, we can imagine that an alien civilization billions of light years away just spotted the explosion of a massive star in our galaxy when it was an infant. Without this luminous event they may have had no idea our distant galaxy existed.

The field containing the explosion seen by Tanvir and Salvaterra is one of six that was observed with HST by Tanvir et al, as described in a 2012 paper in The Astrophysical Journal (arXiv paper). The explosions were traced back to times ranging from about 520 million (*) to about 1.2 billion years after the Big Bang. Only a hint of a galaxy was seen in one of the targets – the nearest one – and nothing at all in the others. Just empty backgrounds. The authors assume that the exploded stars are located in galaxies, because lone stars would be very difficult to understand. Conversely, exotic galaxies are not required. The team estimated that ordinary rates of star formation, at most, are occurring in these unseen galaxies, rates that are about a hundred or more times lower than in z8_GND_5296.

These results reveal an observational bias. The galaxy z8_GND_5296 was only detected at its great distance because it's unusually bright. To use a human analogy, a galaxy like this is over seven feet tall and isn’t representative of the larger population. The indirect detection of galaxies by observing exploded stars also involves limitations: there is an observational bias because the galaxies have to contain regions with young stars, and these destructive events are extremely rare, so huge numbers of galaxies are missed. However, for locating the most typical galaxies in the very distant universe, the exploded star technique still does a better job than direct detection. Tanvir et al. make this point with the title of their paper: “Star formation in the early universe: beyond the tip of the iceberg”.
There are two conclusions I’ll take from these results. One is that results with superlatives are exciting, but they can have limited significance. Another is that non-detections are sometimes interesting, and deserve attention.

(*) Technically this object is more distant but the technique used was not a secure one.
Update (Nov 24th): A few days ago a different stellar explosion was discussed in a NASA press conference and press release. This object, named GRB 130427A, was discovered earlier this year using NASA's Swift Space Telescope and the Fermi Gamma-ray Space Telescope. It was not nearly as distant as the GRBs discussed above, since it occurred at a time when the Universe was a much more mature age of 9.9 billion years, that is about 70% of its current age. However, it has other exceptional properties. The NASA press release, written by Francis Reddy, states that it is one of the brightest GRBs ever seen. Also, a new paper (arXiv paper) led by Alessandro Maselli (one of four papers published in Science) explains that its properties are similar to those of the most luminous GRBs seen in the very distant Universe. In the words of Maselli et al. a "common central engine is responsible for producing GRBs in both the contemporary and the early Universe". The common central engine, in academically understated jargon, is a newly formed black hole (see the illustration below).
In this artist's illustration, the most common type of GRB is shown. The collapse of a massive star forms a black hole (left) and a powerful jet is launched producing radiation across the electromagnetic spectrum. Credit: NASA's Goddard Space Flight Center
Another interesting detail, not mentioned in the press release, is that a supernova explosion was associated with GRB 130427A. Previously supernovas had only been associated with rather weak GRBs observed at relatively close distances. It's now clear that supernova explosions can be associated with very powerful GRBs, like the distant one seen by Tanvir and Salvaterra. This is important because arguments had been made that powerful GRBs didn't have enough power left over for a strong supernova. This result "definitively proves" this isn't the case, as the Maselli et al paper explains.

By coincidence I attended a short talk on GRBs at the Center for Astrophysics on the same day that the Science embargoes went down. The speaker, Tanmoy Laskar, pointed out that there had already been a burst of publicity for GRB 130427A soon after it was discovered, as explained in this slide.

A slide from Tanmoy Laskar's talk. Credit: Tanmoy Laskar
Following Laskar's slide, here's a NASA press release about this "shockingly bright" burst and here is an article by Miriam Kramer from about this explosion being the "most powerful ever seen". This is a good demonstration of the speed that astronomy research can sometimes move at. GRB 130427A was discovered on April 27th, 2013, triggering a bunch of follow-up observations including the all-important estimate of the explosion's distance. The NASA release then followed on May 3rd and press articles soon after that. It wasn't until November 21st that the peer-reviewed science papers were published, confirming the quick analysis performed earlier. For some cosmic perspective on that delay, it took 3.8 billion years for the light from this event to reach us, so six months doesn't seem like much of an extra wait.

Note: In a future, more PR-oriented post I'll use some of these results to look at a few challenges involved with publicizing superlatives.