Author Topic: Thursday 20th October 2011: Supernova Surprises  (Read 6242 times)

Alice

  • Global Moderator
  • Hero Member
  • ****
  • Posts: 31782
    • View Profile
Thursday 20th October 2011: Supernova Surprises
« on: October 21, 2011, 12:50:40 am »
Supernovae have a pretty close place in Galaxy Zoo's heart, with our Supernova project and subforum here, and our "Zoo" logo being inspired by Supernova 1987A.


Wikimedia Commons; also, an old APOD.

They must also have a pretty special place in the hearts of this year's winners of the Nobel Prize for Physics, for their role in discovering dark energy. I've just been doing an assignment for my Astrophysics course which involved a critique of their two papers announcing the acceleration of the expansion of the Universe. You can read Perlmutter et al here and Riess et al here. Riess's was the slightly earlier one. (To download them you can click the PDF button on the top right. It's free.)

To put it as succinctly as possible, the two groups compared a sample of low-redshift, nearby Type 1a supernovae with separate high-redshift (i.e. from z = 0.16 to 0.83) ones - and found that the faraway ones were fainter than they should be, no matter how much they corrected for possible sources of error such as reddening due to dust in space absorbing or scattering light. In other words, there was more space between them and the nearby ones than there should be if the Universe's expansion was going at a constant rate.

(You may remember that I was very confused when I first heard in detail about this result, when I went to Boston. The speaker seemed to be claiming that the supernovae examined where right at the very edge of the visible Universe - they went off when the Universe was extremely young. But, I asked, didn't this mean that there was acceleration going on a long time ago? I spent the next few days getting very unhelpful and often quite angry answers about how expansion works, which basically showed a misunderstanding of my question. This assignment finally fixed that problem. It's revoltingly simple really. The supernovae sample is not at the edge of the observable Universe. All of them are at redshift <1! To a zooite like us, that is not especially far away at all - even though z = 1 is, I am told, about halfway across the visible Universe. No wonder I had been confused, and that also explains why nobody understood my question. This was a great relief. Anyway, forgive that little diatribe, and back to this OOTD . . .)

Type 1a supernovae are excellent standard candles, meaning that they all shine with the same luminosity. This is because they are white dwarfs reaching the Chandrasekhar Limit. This is about 1.4 solar masses. It is the mass at which a white dwarf (or a star's core, if it is not supported by nuclear fusion producing heat and radiation pressure) cannot support itself: its gravity is greater than the repulsion between the electrons in its matter. It becomes a neutron star - alarmingly smaller and releasing a lot of neutrinos as well as a very predictable light curve! Since this always happens at precisely the same mass, it will always generate precisely the same energy.

Or does it?

Although type 1a supernovae are almost perfect standard candles, there are subtle differences between them. One is postulated (though not observed or proved) by this interesting little paper (which is very easy to read and only 4 pages long): we do not know whether a type 1a supernova was from a white dwarf accumulating material from a nearby star, or the merger of two white dwarfs.

This will make no difference to the brightness of the explosion, as it will happen at the same point of mass (even if the two white dwarfs amount to over 1.4 solar masses, the supernova will still occur as soon as the Chandrasekhar limit is reached - all collisions take time). However, it will make a difference to the shape of the explosion and its spectrum. A white dwarf accumulating material (called a "single degenerate" in the paper) is likely to explode in the direction of the star it's shredding, and hydrogen will still be present - even though hydrogen is notoriously hard to detect in spectra (which is why it took until 1925 to discover that it was the most common element in the Universe). But a "double degenerate", the scenario of two white dwarfs merging, predicts a lot more carbon and oxygen, because this is what white dwarfs are made of. The paper also suggests that the less massive white dwarf will be broken up and end up orbiting the more massive one like whatever moon of Saturn's seems to have made its rings! This, too, will affect the shape of the explosion when it happens.

Note that that is a theoretical paper, making some predictions that might explain some differences between Type 1a supernova explosions. It has yet to be proved. They might have got it all wrong. When I last checked (within the last 14 hours . . . just before my assignment was due in :D :D) it had no citations: in other words, no subsequent work has referred to it yet. It will be interesting to find out if they are right!

(Incidentally, you can read about a binary white dwarf system and its fate in this April Object of the Day :) )

Now, what do these two lovely galaxies - a typical red elliptical and a typical blue spiral - have to do with Type 1a supernovae, exactly?


NGC 5080, posted by alandwells, zookeeperKevin,
waveney and Bruno

NGC 3614, posted by lemsgate, greenteagal,
zookeeperKate and mitch

Well, surprisingly enough, the host galaxy seems to have a small but significant effect on the supernova. The literature I was reading for my assignment stated that "late-type galaxies" (i.e. spirals) tend to produce slightly brighter supernovae. ZookeeperKevin sent me this paper about it. Here are a couple of quotes for you to get the idea:

Quote
SN Ia luminosities depend on their light-curve shapes and colours. Using Supernova Legacy Survey (SNLS) and other data, we show that there is an additional dependence on the global characteristics of their host galaxies: events of the same light-curve shape and colour are, on average, 0.08 mag (≃ 4.0σ) brighter in massive host galaxies (presumably metal-rich) and galaxies with low specific star-formation rates (sSFR).

Quote
The observed properties of SNe Ia are known to correlate with the physical parameters defining their host galaxy stellar populations. SNe Ia are more than an order of magnitude more common (per unit stellar mass) in actively star-forming or morphologically late-type galaxies than in passive or elliptical systems (Mannucci et al. 2005; Sullivan et al. 2006). SNe Ia in elliptical or passively evolving systems are also intrinsically fainter, with narrower, faster (or lower “stretch”), light curves (Hamuy et al. 1995, 1996b; Riess et al. 1999; Hamuy et al. 2000; Sullivan et al. 2006).

Well, I got extremely excited when I read this and dashed off an e-mail to Zookeepers Chris and Kevin with a question. Namely: What about red spirals and blue ellipticals?


No catalogue name, just 587736584972271628,
posted by Fluffyporcupine, laihro and waveney.

UGC 9839, posted by DavidWoodward, magsie,
Liz and zookeeper Anze

This was one of the Zoo's excitements in the early days. For those who weren't around then, this blogpost by Steven sums it up very nicely. It was previously thought that a galaxy's morphology - whether it was a spiral or elliptical - determined what colour it was going to be; and what colour it is demonstrates how starforming it is. In other words, spirals are blue and starforming while ellipticals are red and dead. (The blogpost gives the nice analogy of ellipticals being disorganised people living in cities, and spirals as organised people living in the country. Doesn't hold with me personally though. I am a galaxy who's wandered out of the Pembrokeshrie void into the largest supercluster around and have become a lot more organised as a result. But I think this is due to having less time :P :D ;D)

Galaxy Zoo started because Kevin was looking for blue ellipticals. And between all the many thousands of us, we shattered that view. There are red spirals and blue ellipticals, and, by and large, red galaxies live in clusters, while blue galaxies live alone. Star formation takes place where there is free gas, not disrupted or heated or blown off by other galaxies. And a lot of our subsequent work has been looking at other environmental factors: bars, Active Galactic Nuclei, for example, and whether they are driven by or drive colour, etc. etc.

Could supernovae types have anything to do with this, I asked?

I wouldn't have done this Object of the Day at all if either of them had said: "Ooooh! That's a brilliant idea Alice! We must follow this up!" (Visions of being the first to have an idea and getting a Nobel prize, ho hum . . .) Obviously I'm not the first to think of it. Chris told me he's been wondering that for a while - but, as blue ellipticals and red spirals are not especially abundant, we basically don't have enough data to answer it.

But Type 1a supernovae - all supernovae - are objects of fascination. And they are very significant in cosmology, a topic that itself is getting a lot of interest. I am led to believe there are more supernova surveys coming up . . . So stay tuned. Environmental factors among and within galaxies might, or might not, influence them too - even standard candles. Right now - excuse me while I insert this monocle - we Just Don't Know!

zookeeperKevin

  • Galaxy Zoo Keeper
  • Admin
  • Hero Member
  • *****
  • Posts: 2020
    • View Profile
Re: Thursday 20th October 2011: Supernova Surprises
« Reply #1 on: October 21, 2011, 02:55:53 am »
Awesome OOTD!

mitch

  • Hero Member
  • *****
  • Posts: 11116
    • View Profile
Re: Thursday 20th October 2011: Supernova Surprises
« Reply #2 on: October 21, 2011, 03:46:35 am »
Me thinks so too  8) 8) 8) - I don't know where you find the time - impressive....

Budgieye

  • Hero Member
  • *****
  • Posts: 3860
    • View Profile
Re: Thursday 20th October 2011: Supernova Surprises
« Reply #3 on: October 21, 2011, 08:36:54 am »
Lucidly written!
Now I understand more about the binary white dwarf collisions.

graham d

  • Hero Member
  • *****
  • Posts: 2408
    • View Profile
Re: Thursday 20th October 2011: Supernova Surprises
« Reply #4 on: October 21, 2011, 09:54:10 am »
Very nice write up Alice. The origin of those rings still remains enigmatic-
http://hubblesite.org/newscenter/archive/releases/1994/22/text/
and 1987 SN Ia is again topical with regard to the recent and ongoing super luminal? neutrino velocities.












AlexandredOr

  • Hero Member
  • *****
  • Posts: 25423
    • View Profile
Re: Thursday 20th October 2011: Supernova Surprises
« Reply #5 on: October 21, 2011, 10:31:51 am »
Ah yes, great OotD ! Very very very very... informative and pleasant at the same time !
Thx Alice  :D

Baby star opening its eyes on the Universe.

Alice

  • Global Moderator
  • Hero Member
  • ****
  • Posts: 31782
    • View Profile
Re: Thursday 20th October 2011: Supernova Surprises
« Reply #6 on: October 21, 2011, 11:07:51 am »
Thanks everyone ;D

Me thinks so too  8) 8) 8) - I don't know where you find the time - impressive....

Basically by doing it last thing at night, when my head was still buzzing from the assignment after I'd got home from a lecture! I'd eaten, washed up, and had decided I was not doing any more homework (it was already midnight or so) that day . . . and I wasn't going to manage to sleep yet anyway :D :D

If I was more organised I would have done it during the day ;D

Edd

  • Admin
  • Hero Member
  • *****
  • Posts: 2638
    • View Profile
Re: Thursday 20th October 2011: Supernova Surprises
« Reply #7 on: October 21, 2011, 11:16:06 am »
(You may remember that I was very confused when I first heard in detail about this result, when I went to Boston. The speaker seemed to be claiming that the supernovae examined where right at the very edge of the visible Universe - they went off when the Universe was extremely young. But, I asked, didn't this mean that there was acceleration going on a long time ago? I spent the next few days getting very unhelpful and often quite angry answers about how expansion works, which basically showed a misunderstanding of my question. This assignment finally fixed that problem. It's revoltingly simple really. The supernovae sample is not at the edge of the observable Universe. All of them are at redshift <1! To a zooite like us, that is not especially far away at all - even though z = 1 is, I am told, about halfway across the visible Universe. No wonder I had been confused, and that also explains why nobody understood my question. This was a great relief. Anyway, forgive that little diatribe, and back to this OOTD . . .)

I think I understand the problem you were having. Although the universe only accelerates recently - since a redshift of about 0.5 - you need the more distant supernovae to figure this out.

First off, remember that although the universe wasn't accelerating for those very distant z>0.5 supernovae, the light has passed through the universe at times that it was accelerating, and times it wasn't. The light is accumulating information about the expansion of the universe all the time it's travelling - so only for a relatively brief time for nearby supernovae, but for more distant ones that same time, plus the time before that.

Secondly, we're trying to find an acceleration. If you look at a diagram like, say http://www.nature.com/nature/journal/v433/n7023/fig_tab/nature03282_F1.html - that's the difference in the brightness of the supernovae compared to if they were in an empty universe (I know, you can't get supernovae in empty universes... but bear with me!). The three black lines show different cosmological models - the bottom is one with a critical density of matter, the middle with a density of matter 0.3 times the critical, and the top is the one we now generally buy into 0.3 matter and 0.7 cosmological constant. You'll see that all the models match up pretty closely with nearby supernovae - even though that's where the difference in acceleration is biggest. The problem is that nearby you're only looking over a short time. If I show you a photo of a Formula 1 car with a blurry background you can certainly tell it's moving fast, but you can't tell how hard he's accelerating.

Likewise those nearby supernovae only give us a handle on the value of the Hubble parameter - how fast the universe is expanding now. To get a handle on if it's accelerating we need to look back to how fast it was expanding earlier and make a comparison, and see if the Hubble parameter then and the Hubble parameter now have changed in the way we expect.

Does that help a bit more?
When I look up at the night sky and think about the billions of stars out there, I think to myself: I'm amazing. - Peter Serafinowicz

Alice

  • Global Moderator
  • Hero Member
  • ****
  • Posts: 31782
    • View Profile
Re: Thursday 20th October 2011: Supernova Surprises
« Reply #8 on: October 21, 2011, 11:18:12 am »
Thanks Edd! The problem is now completely resolved for me. But your explanation that even if the Universe wasn't expanding then, the fact that it is expanding now means the light has to be stretched on its way, makes even more sense. And I like that about the formula one car too. :)

graham d

  • Hero Member
  • *****
  • Posts: 2408
    • View Profile
Re: Thursday 20th October 2011: Supernova Surprises
« Reply #9 on: October 21, 2011, 03:00:03 pm »
That's an interesting paper Edd from Nature.com. There's no way I can get my hands on the full paper. Take a look at Perlmutter's PhysicsToday paper, Figure 3.
http://supernova.lbl.gov/PhysicsTodayArticle.pdf
There is a distinct excursion at z=0.5- 0.65/0.70  evident in Perlmutter's Fig.4 not represented in Coles figure.
« Last Edit: October 21, 2011, 03:11:57 pm by graham d »

paulrogers

  • Guest
Re: Thursday 20th October 2011: Supernova Surprises
« Reply #10 on: October 21, 2011, 05:39:40 pm »
... the light has passed through the universe at times that it was accelerating, and times it wasn't. The light is accumulating information about the expansion of the universe all the time it's travelling ...

 :D :D :D :D :D
Sometimes we may say something, as if it were nothing--when it really should be trumpeted as the most amazing thing!  Although I'm sure we all know what you meant, this is really one of those amazing things if you think about it.  A massless photon zipping through the Universe at the speed of, well, light, and it's accumulating information as if it were baggage.  Let us never become so jaded that we are incapable of recognizing just how astounding that is!

Edit: It also occurs to me that this is so far outside the common person's experience, even those with a greater than usual appreciation of Science, that we should never "spring it on them", but offer a bit more complete explanation so it doesn't appear so preposterous that Science loses credibility. ;)
« Last Edit: October 21, 2011, 06:06:33 pm by paulrogers »

Alice

  • Global Moderator
  • Hero Member
  • ****
  • Posts: 31782
    • View Profile
Re: Thursday 20th October 2011: Supernova Surprises
« Reply #11 on: October 21, 2011, 06:01:55 pm »
There is nothing quite like the exhiliaration of really appreciating something new, profound, or beautiful, in science. I wrote this because I was buzzing with all this new information about supernovae; how could I go to bed and to sleep? And when I read a book that explained the nature of dark matter to me - that it doesn't lose heat because it doesn't interact with photons, hence, it doesn't condense into solid matter - I was so delighted that I almost leapt out of my train seat and danced. (I contented myself with texting everyone likely to understand instead. :D :D)

Edd

  • Admin
  • Hero Member
  • *****
  • Posts: 2638
    • View Profile
Re: Thursday 20th October 2011: Supernova Surprises
« Reply #12 on: October 21, 2011, 08:48:28 pm »
Sometimes we may say something, as if it were nothing--when it really should be trumpeted as the most amazing thing!  Although I'm sure we all know what you meant, this is really one of those amazing things if you think about it.  A massless photon zipping through the Universe at the speed of, well, light, and it's accumulating information as if it were baggage.  Let us never become so jaded that we are incapable of recognizing just how astounding that is!
That might be exaggerating it a bit. It's more like it's keeping a running total - which isn't quite the same thing. :)
Still, it's amazing how much we figure out from astronomy which is really just collecting lots of photons!
When I look up at the night sky and think about the billions of stars out there, I think to myself: I'm amazing. - Peter Serafinowicz

robert gagliano

  • Sr. Member
  • ****
  • Posts: 422
    • View Profile
Re: Thursday 20th October 2011: Supernova Surprises
« Reply #13 on: October 22, 2011, 03:19:18 am »
Just to muddy up the water, we now have to add to the mix of type Ia SN the 1) super-Chandrasekhar type Ia explosions and 2) the hydrogen-poor superluminous type Ia supernovae and the 3) subluminous type Ia SN with normal width light curve (PTF10ops). Since there are 4 distinct types of WDs....I calculate that there could be 12 (4!/2!) distinct types of WD mergers with different spectra.

« Last Edit: October 22, 2011, 03:45:34 am by robert gagliano »

Alice

  • Global Moderator
  • Hero Member
  • ****
  • Posts: 31782
    • View Profile
Re: Thursday 20th October 2011: Supernova Surprises
« Reply #14 on: October 22, 2011, 06:15:22 pm »
That's not muddying the waters, that's AWESOME! Thank you for sharing! ;D ;D