Eta Carinae's 21-Year Outburst: A Cosmic Instant Replay!

“Not explaining science seems to me perverse. When you’re in love, you want to tell the world.” –Carl Sagan

Nothing lasts forever in this Universe, not even the seemingly timeless stars in the sky. At any moment, any one of the brilliant, twinkling points of light from across the galaxy could run out of fuel, ending its life as we know it. It’s happened a number of times before in recorded history, and will no doubt happen again. With a typical supernova rate of one per galaxy per century, we’ve got a number of nearby potential candidates for what the next supernova to occur in the Milky Way might be.

Today, I’d like to showcase a really special one, and to do that, I’m going to take you far into the southern skies, to the constellation of Carina, the Keel.

Image credit: F. Espenak,

In most constellations, astronomers name the brightest star “Alpha,” the second brightest “Beta,” and so on. So Canopus, the brightest star in Carina and second brightest star in the entire night sky, is also α Carinae, while the second brightest, Miaplacidus, is β Carinae, etc.

Well, almost etc.

For Carina, not only have modern astronomers broken the original constellation up into smaller ones, so that there is now no γ Carinae (it wound up in the constellation Vela, where it’s known as γ Velorum), but something very, very unusual happened to what was, for thousands of years, the seventh brightest star in that region.

Image credit: © 2003 Torsten Bronger, annotated by me.

Where that big red circle lives, with arrows pointing at it, lies what once was the seventh brightest star in the ancient constellation of Argo Navis: Eta (or η) Carinae.

The star is still there, mind you, but it’s not nearly as bright as it used to be. What happened? All was well with the world; there hadn’t been an observed supernova in our galaxy since 1604, when, in 1837, this star underwent a great eruption, becoming much brighter than normal — but not quite as bright as a supernova — for a period of twenty-one years! At its peak brightness in 1843, it was called a supernova impostor, where it temporarily became the second brightest star in the night sky, outshining even Canopus.

Image credit: Celestia, by author / user HeNRyKus.

Since that eruption, η Carinae’s brightness died down so severely that, by the 1860s, it was no longer visible to the naked eye. What exactly happened during that 21-year eruption, from 1837-1858, was a mystery for a very long time. The star wasn’t destroyed; there’s still a Luminous Blue Variable star there to this day. There was also another, minor eruption in 1887, lasting seven years, and it has slowly continued to brighten as time has progressed.

Image credit: University of Minnesota.

So what, exactly, is η Carinae, and what happened here?! One of the most massive stars, weighing it at somewhere around a hundred to 150 solar masses (and somewhere around four million times as luminous as our Sun), η Carinae very clearly underwent some type of eruption. The star itself can be found in a particularly dusty, beautiful region of space known as the Carina Nebula.

Image credit: ESO / Very Large Telescope / T. Preibisch et al., in infrared light.

That’s η Carinae, down at the lower left, surrounded by swirling loops of gas and dust. Spectacular also in visible light, the Hubble Space Telescope got the best view of η Carinae ever back in 1995. Take a look, and see for yourself why, ever since this image, the area around the star has also been known as the Homunculus Nebula.

Image credit: Nathan Smith (University of California, Berkeley), and NASA.

But this was no supernova; the Homunculus Nebula is no supernova remnant and, most importantly, the original star is still intact! You can see it in there, peering out through those explosive clouds, if you zoom deep into the nebula in this very image!

What we think happened, of course, is that just as we get tremors before a massive earthquake, η Carinae had some sort of explosive “hiccup” that will lead up to an eventual supernova. It’s estimated to have blown off about twenty Suns worth of material from its outer layers during this eruption, but the collapse of this star’s core is inevitable. The supernova could come tomorrow or it could not come for another million years; we simply don’t know as much about these ultra-massive stars as we’d like to.

If only we’d had the instruments we have today back in 1837, or even better, in 1843, when η Carinae became a supernova impostor! But we didn’t even have the ability to take photographs back then; all we have are eyewitness accounts from 170 years ago.

But sometimes, the Universe helps us out in ways we could never have predicted.

Video credit: X-ray: NASA/CXC/Rutgers/J.Warren, J.Hughes.; Optical (Light Echo): NOAO/AURA/NSF/Harvard/A.Rest et al.; Optical (LMC): NOAO/AURA/NSF/S.Points, C.Smith & MCELS team.

Above is a video from our satellite galaxy, the Large Magellanic Cloud. From 160,000 light years away, there was a supernova (named SNR 0509-67.5) that occurred about 400 years ago. That is, the light from it reached Earth about 400 years ago; you can see the aftermath here. But, hundreds of light years away was a cloud of gas that reflected the light from the supernova back towards us, giving us a second viewing of that supernova explosion today, hundreds of years later!

This phenomenon is known as a light echo, and it allows us to do something remarkable.

Image credit: Wikimedia user Arkyan.

How is it that we can observe the supernova once again, hundreds of years later? It’s because light can only travel at the speed of light, and the light that takes path B travels a longer distance than path A, while path C is even longer, giving us multiple viewings of the same object, so long as there are clouds of gas for the light to reflect off of. But, unlike hundreds of years ago, we not only have better telescopes, we have photometric filters and spectrographs!

In other words, we can figure out the temperature of the star, what elements are present and in what concentration, and, if we get a light echo, we can watch those things evolve over the course of the explosion!

But getting an echo from a supernova is one thing; getting it from a supernova impostor, because it’s so much dimmer, would be a first.

Image credit: NASA, NOAO, and Armin Rest (STScI) et al.

Welcome to the first supernova impostor light echo ever seen!

In fact, we can learn a tremendous amount about the η Carinae eruption from observations of the echo. From the Hubble press release:

The observations mark the first time astronomers have used spectroscopy to analyze a light echo from a star undergoing powerful recurring eruptions, though they have measured this unique phenomenon around exploding stars called supernovae. Spectroscopy captures a star’s “fingerprints,” providing details about its behavior, including the temperature and speed of the ejected material.

The delayed broadcast is giving astronomers a unique look at the outburst and turning up some surprises. The turbulent star system does not behave like other stars of its class. Eta Carinae is a member of a stellar class called Luminous Blue Variables, large, extremely bright stars that are prone to periodic outbursts. The temperature of the outflow from Eta Carinae’s central region, for example, is about 8,500 degrees Fahrenheit (5,000 Kelvin), which is much cooler than that of other erupting stars. “This star really seems to be an oddball,” Rest said. “Now we have to go back to the models and see what has to change to actually produce what we are measuring.”

Combined with 2003 images from Nathan Smith (who took that picture of η Carinae’s “Homunculus Nebula” above), you can really see the light echo evolve over time.

The full paper details some amazing things we’ve learned from spectroscopy on this light echo, including:

  • The eruption/nebula appears to be expanding at speeds of 210 km/s (!),
  • The star’s eruption temperature is ~5,000 K, much cooler than was previously thought and cooler than the current theoretical models allow for,
  • There are no emission lines, only absorption lines, ruling out the “opaque winds” model, and, in a direct quote from the article,
  • “The cause that triggered such an explosion and the mass-loss without destroying the star is still unknown, but predictions from future radiative transfer simulations trying to explain η Car and its Great Eruption can now be matched to these spectral observations. Other alternative models that were proposed, e.g. the ones that use mass accretion from the companion star… as a trigger for the eruption, can be either verified or dismissed.”

This story isn’t over yet; there are about 15 good years of data about to pour in! And what we find may teach us more than we have any right to know about these ultra-massive, Luminous Blue Variable stars, all because we know to look for a light echo!

(Hat tips to Paul Gilster and Tammy Plotner for their reports on this story, which are also well worth reading!)

18 thoughts on “Eta Carinae's 21-Year Outburst: A Cosmic Instant Replay!

  1. How is it that in 1843, when no one had seen a supernova for over 200 years, there existed criteria to distinguish supernovae from their imposters?

  2. This is what I love about astronomers/astrophysicists. You squeeze your data to extract each and every piece of useful info you can, even if it’s just an echo of something that happened 200 years ago.

    As an applied physicist I very often throw data because it’s not clean enough, or the sample doesn’t look right. You are forced to work with what you have because you can’t do experiments, and I consider that makes you excel at an ability that I always admired.

    Thanks for another great post!

  3. Ethan, Thanks for a great post on news I’ve heard about – your blog helped clarify a lot. Can you tell us more about how they can distinguish the Eta Carinae properties contained in the reflected signal observed, from the properties of the gas the reflection is now coming from?

  4. @Patrick (comment 1) It’s because the brightening lasted 21 years. True Supernovae fade from naked-eye view in a matter of weeks.

  5. As a scientist – how do you react when there’s a 15 year event unfolding? I could take a masters in astrophysics and a PhD on this stuff, crazy!

  6. Question here – not specifically about n Carinae. If this star (or one in our near neighborhood in the Milky Way) were to go supernova, how bright would this appear to us. Is it possible that one within 10-20 light years could be as bright as the moon?

  7. @Phil – my understanding is that no supernova will ever be as bright as our moon. Even the closest star (which isn’t big enough to supernova) would be no brighter than Mars if it could go supernova. The closest star to us (and also to its supernova state) is Betelgeuse, and when it does go supernova, might be like Mars at night (and possibly be able to be seen faintly during the day).

    It’s amazing to me that one can distinguish the reflected light from Eta Carinae’s 1843 explosion from all the other reflected and non-reflected light that’s constantly observed.

  8. Phil @9 / Chris @10,

    The full Moon — on the magnitude scale — has an apparent magnitude of -13 or so (more negative numbers are brighter), while the brightest planet at its most bright is Venus, at around -5. The brightest star, Sirius, is -1, and Mars, which Chris mentioned, ranges from about +2 to -3.

    The brightest supernova ever recorded was the supernova in 1006, which reached a peak brightness of -7.5, brighter than the entire moonless night sky (combined), but not nearly as bright as the full Moon. However, this supernova occurred over 7,000 light years away.

    Betelgeuse is a candidate to be our next supernova; it is only about 640 light years away. If Betelgeuse were to go supernova with the same intrinsic brightness as Sn 1006, there’s your full-Moon brightness.

  9. @ mark. Yes you are the only one

    Follow up question. If betelgeuse were to blow, might it be close enough to cause an electromagnetic disruption? Certainly the particles produced would arrive far after the light

  10. Neat stuff. However, the stellar classification they did is flawed. It uses an algorithm with that was designed to do radial velocity measures for galaxies. Then they exclude from the analysis the parts of the spectrum that would be key to a classification of they type they claim. Its the equivalent of using facial recognition software to determine a person’s ethnicity.. while covering up their eyes and nose. To be fair, no one has been able to do really reliable stellar classification with even high signal to noise spectra. So its a case of over-reaching on an otherwise attention grabbing result.

  11. Must be why I have such an explosive personality, or- on a more positive note, why I’ve been able to achieve great things when I contain and direct my energies positively.

Comments are closed.