Archival Hubble Space Telescope images taken more than a decade ago have revealed an ancient star in the act of exploding.

Five panels are shown. The larger left panel shows the portion of the galaxy cluster Abell 370 where the multiple images of the supernova appeared, which is shown in four panels labelled A through D on the right. These panels show the locations of the multiply imaged host galaxy after a supernova faded and the different colours of the cooling supernova at three different stages in its evolution
Gravitational lensing has revealed the same supernova explosion at three different moments within a single Hubble Space Telescope snapshot. The left panel shows the portion of Abell 370 where the multiple images of the supernova appeared. Panel A shows the locations of the multiply imaged host galaxy after the supernova faded. The image in panel B, taken in 2010, shows the three images of the host galaxy and the supernova at different phases in its evolution. In panel C, the image in Panel B has been subtracted from that in panel A to show three different faces of the evolving supernova. A similar image subtraction shows the cooling supernova's different colors at three different stages in its evolution in panel D.
NASA / ESA / STScI / Wenlei Chen (UMN) / Patrick Kelly (UMN) / Hubble Frontier Fields

Take deep pictures of swaths of night sky using one of mankind's most powerful telescopes, and you never know what you’ll turn up. Such serendipity has led to the discovery a distant star gone supernova just hours before the image was snapped.

In 2010 the Hubble Space Telescope took long images of a galaxy cluster known as Abell 370. The light from this galactic hive comes from 4 billion light-years away. But within the image of this swarm is something else — the light of far more distant background galaxies and the stars within them, their light bent and magnified by the intervening cluster’s mass.

And within that background light are photons coming from a star that “just” exploded — 11.5 billion years ago, when the universe was only 2.3 billion years old.

Gravitational lensing diagram for lensed supernova
The immense gravity of the galaxy cluster Abell 370 acted as a cosmic lens, bending and magnifying the light from the more distant supernova located behind the cluster. The inset at top marks the area where multiple images of the distant supernova are visible. The diagram at bottom shows a magnified version of this area. Thanks to the cluster's warping effect, photons from the supernova in the distant galaxy (shown at bottom right) traveled along three different paths (white lines), arriving at Hubble simultaneously but showing the supernova at slightly different time periods.
NASA / ESA / STScI / Wenlei Chen (UMN) / Patrick Kelly (UMN) / Hubble Frontier Fields

Thanks to relativistic effects of gravity, the chance alignment of this distant star behind the galaxy cluster results in its light being split into four images — an “Einstein cross” alignment. One of these images is too faint to detect; the other three images took different, bent paths to Earth. That means that the three images that are visible capture the supernova at different times.

The photons following the most direct path show the supernova days after collapse. The photons that make up the other two images took slightly longer paths and thus show the exact same event, but 2 days and 8 days earlier, respectively. The earliest image shows the supernova just 6 hours after the stellar core collapsed.

For a star to explode, several things have to happen first. First, the stellar core that has long fought against gravity’s inward pull runs out of fusion fuel and collapses. Regardless of whether it forms a neutron star or squishes further into a black hole, the compact object at the core will meet the outer layers of gas that are still falling in. A reverse shock reverberates outward, eventually breaking through the star’s surface and scattering the outer layer’s gaseous remains.

“What we saw was the emission from after the breakout, when the supernova ejecta expands and cools,” explains study lead Wenlei Chen (University of Minnesota). “Three supernova images show that the expanding surface changed from faint blue (small and hot) to bright red (large and cool).”

These observations reveal the original star’s size. From the temperatures they measured, Chen and colleagues report in Nature that the supernova’s progenitor was a typical red supergiant. Such a size estimate has previously only been possible for a handful of nearby supernovae.

“For me, the most important thing is still the “gee whiz” factor that you can actually find these things and use them to infer properties of supernova progenitors at redshift 3!” says supernova and cosmology expert Robert Kirshner (Harvard University).

While the star that exploded isn’t itself unusual, it’s the second gravitationally lensed supernova to be found in the Frontier Field set of Hubble images, of which this one is a part. (The other supernova, spotted previously, is Supernova Refsdahl.) Using these two supernovae, Chen’s team is able to estimate how many stars are dying in the early universe. The result: when the universe was only a couple billion years old (what astronomers term a redshift of 3), about eight stars would go supernova every 10,000 years within a space 3 million light-years on a side. That number is consistent with our ideas — based on observations of nearer stars — about how stars are born and, ultimately, die.

“As for rates, well, the statistics are very limited: two objects!” Kirshner notes. “Still, this demonstrates the principle, it shows the one previous object was not a fluke, and it opens the path to serious exploration of massive star birth and death at redshift 3. Pretty amazing!”


Tags

Supernovae

Comments


Image of Anthony Barreiro

Anthony Barreiro

November 9, 2022 at 9:18 pm

Wow. Gee whiz indeed. This is really blowing my mind.

How long would this supernova have been visible to Hubble? If Hubble had been observing continuously, would we have a movie of the three lensed images appearing sequentially, each of them brightening and fading and changing color?

You must be logged in to post a comment.

Image of Brian of DRAA

Brian of DRAA

November 11, 2022 at 8:18 am

Betelgeuse, our "local" red supergiant (only ~500 light years away) will shine for about a month when it goes supernova. The three images of this supernova would all fade with the eight day old image fading away first (in say 22 days), then the day 2 image, then the day 0 image. But, at the end of the 30 day observing period, all the info in the day 0 image will have been captured by the day 8 image eight days earlier. Hubble could have stopped "filming" on day 22 and got the whole story. Of course with any SN being so rare, I should think they would have visited this unique target alot if they were aware of it at the time. I think it is pretty cool that the one image provides the brightening and dimming profile of the event at three stages - bang for the buck! Thank you Mr. Einstein.

You must be logged in to post a comment.

Image of Brian of DRAA

Brian of DRAA

November 13, 2022 at 7:30 am

I agree too. I think it was a good catch by RocketMike.

You must be logged in to post a comment.

Image of rocketMike

rocketMike

November 12, 2022 at 1:02 pm

Maybe I'm just confused, but the photograph shows an instant in time that the 3 images arrive at Hubble. Therefore, the short path image is actually showing what happened 8 days AFTER what is shown in the long path image?

***Feel free to ignore my comment if I'm so far off track that it would only confuse the matter more***

You must be logged in to post a comment.

Image of StanR

StanR

November 12, 2022 at 7:03 pm

I agree with @rocketMike; I was thinking the same thing. The photons following the most direct path will arrive at Hubble first; they will thus show the supernova farthest along in its development by the time the photons from the other paths arrive.

You must be logged in to post a comment.

Image of Monica Young

Monica Young

November 14, 2022 at 2:18 pm

rocketMike, good catch! You're absolutely right, and I've fixed this error in the article.

You must be logged in to post a comment.

Image of Lachim

Lachim

November 16, 2022 at 4:12 am

But it still says "The photons following the most direct path show the supernova just 6 hours after the stellar core collapsed.", which is not true.

You must be logged in to post a comment.

Image of Monica Young

Monica Young

November 17, 2022 at 11:01 am

Keeping me honest! That's now fixed too 🙂

You must be logged in to post a comment.

Image of StanR

StanR

November 13, 2022 at 1:25 am

Assuming this is the only photograph during the supernova eruption, how was it determined that the three images correspond to 0, 2, and 8 days?

You must be logged in to post a comment.

Image of Brian of DRAA

Brian of DRAA

November 13, 2022 at 7:52 am

StanR, dandy question. To determine the length of path you would need to "know" or have a good model of the dark matter (DM) distribution of Abell 370 too. I'm guessing the 0, 2 and 8 days are based on the DM model. Then the brightening/fading profile of the SN (super nova) can be matched to known profiles of SN. Then they had good certainty it was a red giant, not massive red giant or type 1A SN etc. I’m just guessing, hope it helps. I used the term DM to mean DM plus baryonic matter. Use of this image will likely be used in DM modelling too (which means they used the DM model to support the SN analysis and the SN analysis to support the DM model... Don’t fall off that stool - you can fall a long way in space, ).

You must be logged in to post a comment.

Image of Monica Young

Monica Young

November 14, 2022 at 2:16 pm

Spot on, Brian. They used a computer model (called GLAFIC, ) of both the early galaxy and the intervening galaxy cluster to determine that there were four images, three of which Hubble detected, and also to determine how much time delay between each image.

I'm not sure what you mean by saying that the image will be used in DM modeling, though. In this paper, the images of the supernova were used to determine what exploded, and then to estimate the rates of those kinds of explosions.

You must be logged in to post a comment.

Image of Brian of DRAA

Brian of DRAA

November 17, 2022 at 10:05 am

Monica, my understanding is, there are a lot of details of the shape of the DM halo yet to be defined. We can't see DM, however, we can see how it effects the path of background light and hence we can infer the shape of the DM halo. The three images of this super nova provide clues as to where the DM resides within the Abell 370 cluster (even the missing image provides a clue). I may have been too keen regarding this image because Abell 370 is “only” 5 billion light years away and I was thinking about how DM figured into the formation of galaxies in the early universe; any super nova, quasar or galaxy that backlights an early DM halo will be gold in refining our model of early galaxy formation. I’m certain Webb will keep us up at night! Cheers, Brian

You must be logged in to post a comment.

Image of rocketMike

rocketMike

November 13, 2022 at 10:11 am

Trigonometry?

You must be logged in to post a comment.

You must be logged in to post a comment.