SN 2018zd: Picturing a New Electron-Capture Supernova in NGC 2146

SN 2018zd exploded in the galaxy NGC 2146 in 2018, and a study published in Nature Astronomy in 2021 argued that it belongs to a class of supernova long predicted and never confirmed: the electron-capture supernova. Turning the result into a press image ran into three problems at once. Hubble had never covered the whole galaxy, it observed the supernova only about a year after the explosion, and the one tool that could have made the star look properly bright would have meant drawing in something the telescope did not record.

A third kind of stellar explosion

Most massive stars die in an iron core-collapse supernova. Over its life the star fuses lighter elements into heavier ones until iron builds up in the core and the star runs out of nuclear fuel. Gravity then wins and the core collapses.

An electron-capture supernova starts differently. In the core of the star, the pressure becomes so extreme that electrons are forced into the nuclei of oxygen, neon and magnesium atoms. Losing those electrons removes the pressure that was holding the core up. The core collapses, and the shock wave tears the star apart.

The result matters for three reasons. It confirms an existing theory with observations, it sheds light on the life cycle of stars, and it may explain the supernova of 1054 AD that produced the Crab Nebula, a long-standing question about that explosion.

Gaps in time and space

NGC 2146 is a barred spiral galaxy in the constellation Camelopardalis, best known for a dusty spiral arm that loops in front of its core, twisted out of shape by the gravity of a neighbor. It is a starburst galaxy, forming stars at a high rate. Hubble had only ever imaged parts of it: an outline of all its observations laid over a Digitized Sky Survey picture shows patchy coverage, with the post-supernova observation a small square near the star.

Timing was the second gap. By the time Hubble pointed at SN 2018zd, about a year after the explosion, the supernova had faded a long way. It is still clearly visible in the Hubble data, but as a modest point of light, not as an explosion that outshone its host galaxy.

Hubble detail on ground-based color

The science team also had color images from the Las Cumbres Observatory taken close to the supernova's peak. There SN 2018zd is one of the brightest objects in the frame. The ground-based data, however, are blurred by Earth's atmosphere, so they cannot simply be pasted together with Hubble's sharper view.

Hubble's color information was thin: the galaxy had been observed in only two quite different filters. The solution was to treat the two datasets as two halves of one image. The Hubble F814W data, combined from two observations, became a luminosity layer, the lightness of the image, placed over the Las Cumbres image, which supplied the color. Human vision is very good at merging low-resolution color with high-resolution lightness, and the result is perceived as one sharp, fully colored picture. It worked here because the Las Cumbres color data cover wavelengths similar to those of Hubble's red F814W filter.

The second Hubble filter, F658N, was added too. It isolates light from hydrogen and nitrogen and picks out the star-forming regions as bright red patches, most visibly toward the top of the galaxy. The colors follow the usual conventions described in Are the Colors Real?.

The synthetic star that was left out

That left the supernova itself, still dim in the Hubble layer. One experiment used Tiny Tim, a program that models how Hubble's optics spread the light of a point source, to generate a bright star with exactly the luminosity of SN 2018zd at peak and exactly the instrument setup of the Hubble observations. It looked convincing.

It was rejected. A computed star would have been an element not derived from the data, however close to reality. Synthetic stars like this are acceptable in a 3D fly-through, where there is no choice but to represent unseen depth artistically, but a press image of a scientific result has to be built from what was observed.

The compromise blended the Las Cumbres observation of the supernova at peak into the Hubble image. The star comes out large and soft, far from Hubble's usual pinpoint sharpness, but every pixel of it is real data, and it shows what the study is about: a dying star bright enough to outshine its galaxy. The same rule, use real data first, governs routine cleanup too, as described in Image Artifacts Explained.

Questions readers ask

What is an electron-capture supernova?

It is a supernova in which extreme pressure in a star's core forces electrons into the nuclei of oxygen, neon and magnesium atoms. Without those electrons the core can no longer resist gravity, collapses and triggers an explosion. It had been theorized long before SN 2018zd was found.

Is SN 2018zd related to the Crab Nebula?

Indirectly. The Crab Nebula is the remnant of a supernova seen in 1054 AD. Confirming that electron-capture supernovae exist gives astronomers a new way to look at the origin of that explosion, which may have been of the same type.

Why does the supernova look blurry in the press image?

The bright supernova came from a ground-based image taken by the Las Cumbres Observatory near peak brightness, blurred by Earth's atmosphere. Hubble observed it only a year later, when it had faded, and a computer-generated sharp star was rejected because it would not be real data.

What is a luminosity layer?

It is an image used only for the lightness of a picture, while another image provides the color. Placing sharp Hubble data as luminosity over lower-resolution color data produces a picture that looks both detailed and colorful, because human vision relies on lightness to judge sharpness.