Happy 25th annniversary, Supernova 1987A!

By Phil Plait | February 23, 2012 7:01 am

[Note: This is a repost of what I wrote on February 23, 2007, the 20th anniversary of the day light reached us from Supernova 1987A, the brightest supernova that had been seen for centuries. Not a whole lot has changed since I posted this, including my opinions on it, so I figure I’ll give it a second chance (I kept all the dates the same as they were five years ago, so be aware of that as you read it). That exploding star 1.6 quintillion kilometers away certainly changed my life – I could even say I got married because of it, though that’s a story for a later time – and it’s all been pretty good, I must say.]


Has it really been 20 years since Sanduleak -69 202 blew up? Wow.

Of course, that’s Supernova 1987A to you. And if that doesn’t ring a bell, this image might:

That image, from Hubble, was released as part of the 20th anniversary of the closest supernova in 400 years. It’s amazing– we know so much about this event now, and I know a lot of people — including me — who would have killed for this knowledge back in, say, 1990. But we still don’t understand a lot about it too.

If you want a history of this object, I won’t belabor it here, since I have details in an article I wrote about 87A. In fact, that is the first part of a series of short articles I wrote about the supernova (at the end of each is a link to the next). But the early days were very confusing. It was thought only red supergiant stars could explode, but this one — named Sanduleak -69 202 — was clearly blue when it detonated. It emitted ultraviolet light as expected, but the amount was weird– it got brighter and fainter in an odd way, as if it had a cloud of gas around it. The best images we could do, pre-Hubble, did show some sort of elliptical envelope of gas, but the UV light didn’t match the shape seen.

It wasn’t until we got the Hubble images in 1990 that we saw that inner ring, and things made a little more sense. We could see that it wasn’t a complete shell (like a football), it was a flat ring! But then we realized we still had a problem: how did that ring get there? And what were those fainter arcs above and below it?

I was in the thick of it back then. I had just started my PhD research, signing on with a large project to look at exploding stars with Hubble. I signed up just before launch, so I got to live through the trauma of learning about the flawed mirror, and getting the fuzzy data. I spent months learning how to clean up the data, and wishing for just a few more photons, and lying awake at night (after getting our first meager data on 87A) staring at the ceiling trying to figure out just what the heck that ring was.

Lots of false starts. Lots of dead ends. Lots of great ideas smashed by reality. Lots of math. Lots more math. Lots of meetings, lots of talking, lots of sitting in front of a computer learning about deconvolution, pointing constraints, spectral analysis, Fortran, IDL, Unix.

In the end, I was able to cobble enough together to write a scientific paper and get my PhD in 1994. A lot of what I (and my collaborators of course) were able to figure out turned out to be right, and some turned out not to be. No matter how you slice it, Supernova 1987A is a weird object. For a long time we knew of nothing else like it, but eventually (really, quite recently) people found similar objects. Not that we understand how you can get those three rings like that– that’s still a mystery, even after 20 years.

Wow, 20 years. Well, the Universe ticks on. During that time, the inner ring faded as the initial blast of light from the explosion wore down, but then it rebrightened years later as gazillions of tons of hot gas from the exploded star begin to reach the ring. The supernova itself was at first an unresolved dot, but as you can see from the image above it’s expanded greatly over two decades. It’s an elongated cylinder of material now, getting bigger every day.

And it won’t stop, not for thousands of years. After 7305 days, it’s stretched out to be a goodly fraction of a light year, but it’s still screaming along at thousands of kilometers per second. It’ll slam into the inner ring, eventually dispersing it (that’ll take decades, probably). A few hundred years later it’ll reach the outer rings, and blast them apart too. Maybe eventually it’ll look like the Crab Nebula; there’re some indications the explosions were a bit similar. Even then it’ll go on, getting bigger and fainter, looking like the Vela remnant, and then Simeis 147, and then it’ll fade, mix, and merge into the ambient gas surrounding the other stars in the outskirts of the Tarantula nebula, where the star that made the supernova was born. No doubt by then more of the stars in its neighborhood will blow up; the place is lousy with massive stars just waiting to go off. They’ll have their turn, and future astronomers can marvel over them, as well.

I hope they have as much angst, sleepless nights, head-scratching, wonder, joy, awe, and fun as I did looking at Supernova 1987A. Isn’t that the point?

Note added after I edited this, but before I posted it: Amazingly, while looking up some info about the star that exploded, I stumbled on the announcement that the mystery of the origin of the three rings may be solved! It’s been speculated for a while that the star that blew up was originally a binary star, two stars orbiting each other. If one was more massive than the other, then it could have literally swallowed the smaller one up when it expanded into a red supergiant (it turned blue later). The smaller star spiraled into the bigger one, eventually reaching and merging with the more massive star’s core. As it spiralled in, it "spun up" the more massive star, making it rotate faster and flattening the equatorial regions into a disk. That’s how the inner disk may have formed. Eventually, much of the outer gas of the merged stars was ejected in various stages, and the complicated ejection mechanism may have formed the outer rings. New models by Podsiadlowski, Morris, and Ivanova appear to confirm this idea mathematically, which is fantastic news! It’s not 100% certain, of course, but it’s a great step toward understanding. And that, most certainly, is the point.

Update (late on February 23): the wonderful and talented Jennifer Ouellette, whose name I cannot type correctly on the first try no matter how much I want to, has much more meaty info on this.

Image criedt: NASA, ESA, P. Challis & R. Kirshner (Harvard-Smithsonian Center for Astrophysics); Hubble & Plait

CATEGORIZED UNDER: Astronomy, Piece of mind, Science
MORE ABOUT: Hubble, Supernova 1987a

Comments (27)

  1. Messier Tidy Upper

    Happy anniversary SN1987 A indeed. :-)

    Now if only we could arrange some further stellar fireworks to celebrate the occassion! 8)

    Or find that missing central pulsar – or whatever remnant is left of the core of Sanduleak -69 202 – at last as a present! :-)

  2. Messier Tidy Upper

    Pretty overtired right now so I might be missing something basic here but :

    The smaller star spiraled into the bigger one, eventually reaching and merging with the more massive star’s core.

    Could this merger have triggered the supernova and made it different, perhaps explaining the apparent lack of a stellar remnant? Has that hypothesis been advanced for this yet?

    SN1987 was unusual in lot of ways – blue supergiant star rather than red variety exploding thus under-luminous absolute magnitude and different light curve, lack of predicted remnant pulsar / black hole so I gather and more.

    What effects – other than turning a red supergiant blue – might that stellar merger have had on the course and outcome of the 1987 a supernova?

    Wondering when our Milky Way Galaxy will next witness a supernova – hoping its Eta Carinae in my lifetime, soon~ish!

  3. TM

    “Has it really been 20 years since Sanduleak -69 202 blew up?”

    Actually, hasn’t it been 168,020 years? Although I guess that depends on your frame of reference. :-)

  4. Cindy

    Happy anniversary SN1987A.

    Thanks for the trip down memory lane, Phil. Is there any new information, theories, etc. about SN1987A in the past 5 years?

  5. Arie

    So Wikipedia gives the 24th as a date. Here and on the linked site, 23rd is used… Which is it?

  6. abadidea

    “learning about deconvolution, pointing constraints, spectral analysis, Fortran, IDL, Unix.”

    As a computer scientist, I am so, SO sorry you ever had to deal with that monstrosity that is IDL.

    I literally dropped the book in fright several times when reading about it. And then raged. And felt a little more dead than before I had read about it.

    (Note: there are two IDLs, I assume he means the one oft used in astronomy.)

  7. That top picture is just soo pretty. I can see why you wanted to repost this! And the things we learn from events like this just makes it all the more satisfying. :)

  8. Chris

    @3 TM
    I don’t know the exact error bars on the distance, but I’m guessing it’s probably between 100-1000 light years, so it could be impossible to pinpoint it’s exact supernova date exactly. Maybe someone else has an idea on typical distance errors.

  9. Jeff

    I’m a half glass empty kind of guy, which is why I’ve been a prof so long, the kids kick me in the butt to cheer up , at least when I’m around them.

    I see lost opportunity here for 90% of the people north of equator, who could have seen this; but it teased us and now I doubt I’ll see a bright supernova. Would have loved to have seen it, but at least at the time we had a lot to talk about it. At least we got Hale-Bopp as a spectacular northern hemisphere event in our lives.

  10. Chris

    @9 Jeff
    Don’t forget those “Southerners” had two really great comets. McNaught and Lovejoy. I’ve got to move!

  11. amphiox

    SN1987 was unusual in lot of ways

    What are the odds that the first supernova we got to see that was nearby and that we really got to examine in detail with modern astronomy instruments would turn out to be an unusual one, rather than an average one?

    Could it be that SN1987A was not actually unusual, but just appears unusual to us due to insufficient knowledge about the true diversity of supernovas?

    And how would this consideration apply to probability statements founded on similar logic, such as “well, this earth-like planet is only 20 light years away, and so for us to find one now, so soon after looking, must mean that earth-like planets are very common in the galaxy”?

  12. MarcusBailius

    So it’s the 25th anniversary of the first measurement of the speed of neutrinos, confirming it was less than the speed of light…
    :-)

  13. OneofNone

    @5. Arie: Good question.

    Which Wikipedia do you check?
    German Wikipeda says “24. 2. 1987 (23:00 UTC)”.
    English Wikipedia has the same time in the table at the right. But in the first paragraph February 23rd is given. Later it is the 24th. That’s definitely strange.

    At 23:00 UTC some regions on Earth already had Feb. 25th as date, but as far as I know the 23rd had passed on all locations.

  14. OneofNone

    @12. MarcusBailius:

    Unfortunately there is no proof of the Neutrino speed in any favour. They arrived before the visible light by about 3 hours. So they have been faster.

    On the other hand the light is thought to have been halted for some time. Sounds like a loose fiber, doesn’t it? :-)

  15. Dutch Railroader

    @amphiox – No, SN1987A is unusual in a number of its gross properties. Both before and after we have seen many Type II SN, and can compare their luminosity and time evolution to 87A. The proximity of 87A provided many subtle details, such as the SN progenitor star and the core-collapse neutrinos, as well as a rich look at the SN environment and the interaction of the SN with it. But again, we have many SN to compare to its overall behavior.

  16. Dutch Railroader

    @12 & @14

    SN 1987A is some ~170,000 light years away, so a few hour time difference is a trivial difference in velocity as compared to the CERN implied difference in neutrino and photon speed (which would have amounted to a 3 year difference or so if scaled to the SN 87A distance). This really can’t be pushed further. The neutrinos mark the exact time of core collapse, the triggering event of a SN, to within a second or so. The resultant shock wave that creates the SN light has to climb out from the core of the star to the surface, which takes a while (I’m not an SN expert, so I forget how long this is). The initial brightness rise for SN 87A was not observed, so its timing has to be inferred by extrapolating the light curve back in time. SN 87A was caught well before its maximum brightness, so it was possible to do this with some accuracy, but given the errors, it’s never going to be better than a few hours…

  17. Allen Thomson

    A question about the bright white spots around the reddish ring: They appear to be somewhat regularly distributed. If you Fourier analyzed their distribution, I’d guess there’d be a fairly sharp spike. Is that true and, if so, why?

  18. Messier Tidy Upper

    @11. amphiox :

    What are the odds that the first supernova we got to see that was nearby and that we really got to examine in detail with modern astronomy instruments would turn out to be an unusual one, rather than an average one? Could it be that SN1987A was not actually unusual, but just appears unusual to us due to insufficient knowledge about the true diversity of supernovas? And how would this consideration apply to probability statements founded on similar logic, such as “well, this earth-like planet is only 20 light years away, and so for us to find one now, so soon after looking, must mean that earth-like planets are very common in the galaxy”?

    Well co-incidence does play a role and the sample size is far too small.

    For instance, yellow dwarf stars similar to our Sun make up only around 4% of all stars with nearly all being much cooler, fainter red, white* and (if you call ‘em stars*) brown dwarfs. Yet Alpha Centauri – our nearest stellar neighbour – contains two such stars albeit Alpha Centauri B is an relatively bright, hot and massive example of an orange dwarf.

    Toss a dice five times in a row and you’ll get lucky and throw five sixes on occassion – but that won’t be a probable result. Unless the dice is weighted natch! ;-)

    Only when the sample size grows eg. we’ve made a thorough and complete as possible survey of all the nearby stars and their exoplanets or have a larger list of supernovae do we get a real idea of how typical / likely such things are.

    ——————————————————————-

    * Technically speaking you could also argue that white dwarfs are no longer “true” stars but stellar corpses / remnants – although few people do.

  19. OneofNone

    @5. Arie:

    I guess I found the answer to your question. According to The Ultimate Neutrino Page the Neutrinos have been observed on February 23 7:35 UTC. So this is taken as the time SN1987A happened (the Core collapsed).
    The light has been seen first on February 24 23:00 UTC. That’s after SN1987A became visible. The light has been there earlier, but someone had to notice that.

    As Dutch Railroader explained in #17, the visible event really happened at some time in this window of 39:25 hours.

  20. Daniele Fargion

    The SN 1987 A had many merit:
    nearest SN explosion, first neutrinos, bounds on neutrino masses, first hint of axialsymmetric SN (jet like) collimated but off axis as most SN-GRB jet events.

    In my opinion the ring and the nebula inside are showing an hour-glass structure soon or later revealed as a precessing and spinning jet. DF and A Salis
    inAstrophysics and Space Science
    Volume 231, Numbers 1-2, 191-194, DOI: 10.1007/BF00658614
    Precessing Gamma jets, GRB and the twin rings around SN1987A ..

    However the SN1987A neutrino event did help (to some of us) to severely constrains and reject the so called (embarassing or unfortunate ) recent super-luminal claim (Opera-Cern) on neutrino tachyon speed that we now feel coming back to reality…see arXiv:1109.5368 .
    Best regards DF

  21. Messier Tidy Upper

    @ 2.

    Pretty overtired right now so I might be missing something basic here but could this [stellar merger theory – ed] have triggered the supernova and made it different, perhaps explaining the apparent lack of a stellar remnant? Has that hypothesis been advanced for this yet?
    What effects – other than turning a red supergiant blue – might that stellar merger have had on the course and outcome of the 1987 a supernova?

    On further reflection I guess the time scale for that – millions or at least thousands of years earlier – rules it out as the trigger, probably?

    Although I’d still love to see the second question there answered if anyone is still reading and can enlighten us.

  22. Matt B.

    @0. The link on “Simeis 147″ ends up going to the current BA article. It must have a misspelling or something. But this works: http://blogs.discovermagazine.com/badastronomy/2005/11/13/sky-factory/

NEW ON DISCOVER
OPEN
CITIZEN SCIENCE
ADVERTISEMENT

Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

ADVERTISEMENT

See More

ADVERTISEMENT
Collapse bottom bar
+

Login to your Account

X
E-mail address:
Password:
Remember me
Forgot your password?
No problem. Click here to have it e-mailed to you.

Not Registered Yet?

Register now for FREE. Registration only takes a few minutes to complete. Register now »