Getting a handle on dark energy is one of the great goals of modern physics. But understanding what it is that seems to be accelerating the expansion of the universe depends upon the accuracy of our measurements. We can study this acceleration by looking at the behavior of Type Ia supernovae, which can be used as ‘standard candles’ — the distance to a galaxy can be measured because the visual magnitude of this type of supernova depends on its distance. But how reliable are our standard candles? New work confirms the usefulness of these stellar events while explaining why some supernovae can look different from others.
A Type Ia supernova occurs when a white dwarf gathers material from a nearby companion star and approaches the Chandrasekhar limit (about 1.38 solar masses), at which point the pressure and density have grown beyond the point that the star can support its own weight. Various processes have been invoked to explain the details, but while these supernovae seem alike, some recent studies have shown that the speed with which their spectral features evolve can vary. This change in the so-called ‘velocity gradient’ was noticed around the same time that Type Ia supernovae were first being used to flag the existence of dark energy.
Image: A schematic picture of structure of type Ia supernovae derived by the observations. The ash of the initial sparks is at an offset (yellow). Depending on the viewing direction from which a supernova is observed, a supernova manifests different spectral properties: If viewed from the offset direction, it appears as a “Low-Velocity-Gradient” SN and shows blueshift in late-times. From the opposite direction, a SN is a “High-Velocity-Gradient” SN and shows redshift (as is shown in the observational data in Fig. 1). Here, low- and high-velocity-gradients SNe are those showing a slow and rapid evolution in spectral features, respectively. Credit: IPMU/University of Tokyo.
By relating the velocity gradient to the wavelength shift of emission lines in spectra taken some 200 days after a supernova explosion, Keiichi Maeda (University of Tokyo) and team have found that the degree of the gradient depends on the direction from which the supernova is observed. These supernovae explode asymmetrically, the team argues, and because the events are observed from random directions, the differences will average out. Says Maeda:
“This is a single stone to kill at least three birds. It is not only about uncovering the origin of the spectral diversity. The finding now gets rid of a concern about using Type Ia supernovae for cosmology, since this viewing angle effect will average out if we collect many supernovae for cosmology. In addition, the idea of the uniform progenitor system is rescued. Finally, this is the first strong observational indication about how the thermonuclear flames are ignited in the explosion — the finding points to asymmetric, off-center explosions, as opposed to what most people had believed so far.”
Thus the critical supernovae are still standard candles despite variations in their spectral features, confirming the validity of the measurements of dark energy produced by the High-z Supernova Search Team and subsequent researchers. We still have a robust mystery on our hands: What is it that seems to be acting against the gravitational force to accelerate an already expanding universe? We’re not through with Type Ia supernovae yet — we still have several supernovae whose stars seem to have expanded beyond the Chandrasekhar limit before the explosion occurred — but learning that the great bulk of these events are still reliable indicators is strong support for the dark energy they have uncovered.
The paper is Maeda et all, “An asymmetric explosion as the origin of spectral evolution diversity in type Ia supernovae,” Nature 466 (1 July 2010), pp. 82-85 (abstract). Also available: Kasen, “Astrophysics: The supernova has two faces,” in the same issue, pp. 37-38.
Here’s a video of a simulated Type Ia Supernova. It’s very interesting — and I believe it illustrates the asymmetric ignition phenomenon discussed here. (I believe it was produced by the Advanced Simulation and Computing (ASC) Academic Strategic Alliances Program (ASAP) Center at the University of Chicago — linked here.)
Excellent article. Now that we can reassert our confidence in the SNIa luminosities, we can refocus our attention on the real problem of the “Dark Energy” mystery: false interpretation of redshifts.
As far as I can see, nothing in this story adds or subtracts anything to the proof or otherwise of dark energy existence.
“Grey extinction” also explains the standard candles observations without recourse to dark energy. Models have been fitted where the concentration of the “grey extinctor” increases with increasing red shift (z) in a cubic relation, as may be expected if the universe is expanding.
Does this address the question as to whether these events are more often caused by accretion or from merger events? If merger events make up a substantial proportion of the SNIa then there may still be issues with using them as standard candles, as there is no reason for a WD+WD binary to have a mass equal to the Chandrasekhar limit.
kzb: I do not believe that the “extinction” idea has any traction. The amount of intergalactic material required would have to be implausibly dense. You’d then be forced to reinvent dark energy anyway — just to explain why the additional mass did not halt cosmic expansion eons ago.
andy: I do not believe that merger events make up a substantial proportion of the SNIa data. The idea gets a lot of airplay, since it poses a challenge to standard candles. But star collisions in galaxies are exceedingly rare. The best odds for collisions occur in elliptical galaxies (where stellar orbits are generally random — as a opposed to spiral galaxies, where stars are generally co-rotating) but my understanding is that most SN1a datapoints come from spiral hosts and that the datapoints from elliptical hosts are not out of line.
Erik Anderson: The merger scenario does not rely on the collision of two independent stars (which would indeed be rare), but the coalescence of a WD+WD binary, which is a far more likely event. Although according to this paper, such supernovae would be fainter than the accretion-produced SNIa…
To andy: Ah — indeed you said ‘binary’ up above. I can’t think of a mechanism that would relate binarity to galaxy-distance.
Erik Anderson: not if the “grey extinctor” also replaces some or all non-baryonic dark matter?
http://arxiv.org/PS_cache/arxiv/pdf/1003/1003.0453v1.pdf
To kzb: The purported whereabouts of “cold dark matter” are concentrated in galaxies and galactic clusters — not strewn throughout the cosmos.
Erik Anderson: Schild’s theory is that there are indeed clumps of frozen hydrogen/helium floating around in vast numbers. These clumps have become coated with dust because they have been floating around since the universe condensed. When they are heated, a lot of stuff is driven off, much like a comet but in interstellar space. He contends that this explains the supernovae observations at high z.
Erik Anderson: well it is known that metallicity does have an effect on stellar evolution, I guess this may have implications for the kinds of systems formed if mass transfer is involved, and given that metallicity is expected to increase with time you’d presumably get some variation in the populations of WD+WD binaries. Very handwavey I know, not saying I agree with this explanation. But it might be an effect that should be factored in, it might turn out to be negligible.
It appears that headway is being made concerning understanding of type 1a supernova but there is now a seemingly good competing theory concerning collisions of two binary white dwarf stars. There may still be a lot we don’t really understand about them but the bottom line I think is that there is little doubt that they are standard candles of a sorts destined to be an invaluable future distance indicator which will be very interesting as time goes on concerning redshift distances vs. the luminosity distances of supernova.
As to their being evidence of dark energy I believe this will be overturned in the next 5 to 10 years. Their divergence from the empty universe/ perfect standard candle line I think will be explained by changes to the Hubble distance formula which I think will end the dark matter hypothesis but strengthen type 1a supernova as standard candles somewhat correcting calculated distances, maybe about 10% in the foreground would do it.