Some say the world will end in fire,
Some say in ice.
From what I’ve tasted of desire
I hold with those who favor fire.
But if it had to perish twice,
I think I know enough of hate
To say that for destruction ice
Is also great
And would suffice.” –
Robert Frost

We’ve come to believe that we know the fate of the Universe: that dark energy will drive distant galaxies and clusters apart, leaving only the objects that are already gravitationally bound together intact.

Image credit: NASA/JPL-Caltech.

Image credit: NASA/JPL-Caltech.

Through the use of type Ia supernovae as standard candles, we’ve been able to trace out the Universe’s expansion to tens of billions of light years, and the big surprise was that the expansion is speeding up, and hence we’re in for a terribly lonely long-term fate.

But what if these supernovae weren’t so standard? A new discovery may indicate exactly that.

Image credit: screenshot from

Image credit: screenshot from

Does this mean that dark energy might be all wrong? Find out on this week’s Ask Ethan!


  1. #1 David
    April 11, 2015

    Thank you again, Ethan, for another great explanation of an important topic. Love your blog.

  2. #2 Omega Centauri
    April 11, 2015

    So are we able to tell from some combination of spectra and/or lightcurve without knowing the distance which population a given SN belongs to?
    Also I would imagine the net mass of the merger-1a’s varies quite a bit, and this (and possibly other factors such as the mass ratio or composition) might make this class more variable.
    A second claim, was that the ratio of the two populations of 1as might be time dependent.

  3. #3 João Carlos
    Rio de Janeiro, Brazil
    April 11, 2015

    One point that catcehd my attention is Milne’s statement: “As you’re going back in time, we see a change in the supernovae population (…)”
    As you pointed out, tipe Ia supernovas have two different “gestations”. I’ll call them “accretion” and “merger”. It seems plausible to me that accretion was far more common in the early universe.
    Do you see any possible correlation?

  4. #4 Mamobu Mabunda
    Jane Furse
    April 11, 2015

    I find this blog very interesting, what intrigues me the most is that we can actually use light to determine which population belongs to which supernova, thank you for sharing with us this great insight

  5. #5 Wow
    April 12, 2015

    Re 3: Is there any reason to expect your hypothesis to be true? There are a billion wrong ideas out there, what makes you think that yours isn’t one of them?

    That’s the start of scientific skepticism.

  6. #6 kevin
    naples florida
    April 12, 2015

    the revelation that our galaxy and probably most others are at least 50% larger than previously believed, will certainly have an effect on how much dark MATTER is required to keep a galaxy in tact, and by extention, how much is believed to be present in the universe… added to this..revelation on dark ENERGY.. is an exciting time for cosmology and science in general….
    its great how science LOVES to find out it was wrong.. so unlike religion.

  7. #7 Khnayisile Hlongwane
    April 12, 2015

    It is interesting to see how a small object like a candle can be used to determine such a huge thing about our universe.The fate of the earth in terms of how it will end might not be known at present but,with new discoveries made that mystery might be solved sooner then we expected.(u15097367)

  8. #8 Boris Borcic
    April 12, 2015

    Do (asto-)physicists maintain an accurate mind model of how particular measurements may be critical, but on a non-proximal way, to particular conclusions – like in anticipation of potential retractions? To outsiders like me, there remains doubt that the supplementary evidence for dark energy after the initial SN 1a studies might hide that it is less autonomous that overtly appears, and a correlative suspicion that a later revolution in understanding could thus involve less of a miracle than this summary suggests. There is something special to this form of doubt, it has an ingredient of Kant’s imperative to it, on the one hand you’d prefer not to expect smart experts capable of overlooking that, but on the other hand the easiest way for all to overlook such things would be for all to expect smart better experts to take care of the worry.

  9. #9 Michael Richmond
    April 12, 2015

    You can read a copy of the full technical article on which this post is based at

  10. #10 Sigciniwe
    South Africa
    April 12, 2015

    Wow! brilliant work! Are there possible ways to solve this? How many years is this going to take?

  11. #11 Sigciniwe
    South Africa
    April 12, 2015

    Wow! Brilliant work! Are there possible ways to solve this? How many years is this going to take?


  12. #12 Pule
    South africa(Pta)
    April 12, 2015

    Supernovae can explode in galaxies, and the are used to trace large-scale structure of the universe. Which helps astronomers to trace the expansion of the Universe.

  13. #13 Podile Seromo
    April 13, 2015

    If the universe is expanding to tens of billionth of light years will in some point this expansion stops?

  14. #14 Ben
    April 13, 2015

    What are the implications of these discoveries on the ultimate fate of the universe, if any?

  15. #15 Sal Minella
    New York
    April 13, 2015

    If a galaxy one billion light years away appears to be moving away from us faster than a galaxy one million light years away, it would seem that the universe was expanding faster one billion years ago than it was one million years ago. If that were the case then, it would follow that the rate of expansion of the universe is slowing.

    How is this logic contradicted as it seems to be every time that I read that more distant objects are moving away from us faster than less distant objects?

  16. #16 Sean T
    April 13, 2015

    Sal Minella,

    The answer must take into account the quantitative rates of expansion. To see this, assume that the rate of expansion has been the same for all times. Given that assumption, the more distant the object you observe, the more rapidly you would see that object move away.

    To see this, assume for simplicity that each unit of space will double in size during each unit of time. That defines a constant rate of expansion. Now, consider object A located one unit of space away. After one unit of time, you would observe object A at a distance of two units of space. It has moved with an apparent speed of 1 space unit per time unit.

    Now consider object B, located 2 units of space away. In one unit of time, each of those 2 space units doubles in size. Therefore, after one time unit, we observe object B at a distance of 4 units of space. Object B has apparently moved 2 space units in one time unit, implying a velocity of 2 units of space per time unit, which is twice the apparent velocity of object A.

    Therefore, for a constant expansion rate, we see that more distant objects are expected to appear to recede more rapidly than closer objects. It is deviations from that expectation that inform us as to whether the rate of expansion is increasing or decreasing over time.

  17. #17 Sal Minella
    New York
    April 13, 2015


    Thanks for the response. Consider me a simple man with a tenuous grasp of relativistic concepts. As I understand your explanation, space-time could be envisioned as a rubber sheet that is being stretched from it’s outer circumference.

    This does not really address the issue that is confounding me – that is that any EM radiation that we receive from a given distant object is essentially as old as the distance of that object divided by the speed of light (ideally).

    Given that, when we evaluate that EM radiation here on Earth we are evaluating the EM that was radiated at the previously calculated time. Given that, when we compare the redshift of two distant objects A (one million light-years away) and B (one billion light-years away) are we not actually comparing the redshift from two different times relative to us? If that is the case then one cannot draw any conclusions about what objects A and B are doing right now (here and now Earth time and space) of, for that matter, if objects A and B still exist.

  18. #18 Russ
    April 13, 2015

    Hi Ethan,

    Thanks again for a great article.

    At one point you say “Are we sure there isn’t some new type of dust or some other light-dimming property (like photon-axion oscillations) at work here? As it turns out, these issues were all able to be resolved and ruled out; these things aren’t issues.”

    Could you expand on what light-dimming things have been considered, and what experiments have been run? It’s something I’ve always wondered about.


  19. #19 Wow
    April 13, 2015

    Sal, you’ve changed your question now that you’ve been answered by Sean, to something completely new. Is Sean going to be playing whack-a-mole with your problems in helping you out?

    There is nothing in your new statement that accords with anything mentioned in your first. This does not appear to be fair on Sean.

    The answer to this one is not easy without dealing with the mathematics of spacetime metric. The hand-waving version is that since the photon is perusing the same event, the time of when it left is noted by when we receive it. Even if the source no longer exists by now, it doesn’t matter. This no more makes the photon weird than the fact that your mother will (hopefully) die some time in the distant future, but you will still go on living, despite her being the source of you.

    The redshift is appropriate to the source at the time the photon was emitted, no matter its’ state now.

  20. #20 Sal Minella
    April 14, 2015

    Thanks for your response. The problem with my question may be that it is too simple to answer. Restated: If we are comparing the redshift of two objects, what conclusions can we draw about their relative positions when we really don’t know where they are now or even if they exist now? The question is not what do those photons mean but, given the “when” of their creation, how do they relate to one another.
    BTW these are not gotcha questions, just genuine curiosity,

  21. #21 Sean T
    April 14, 2015


    Yes, while analogies are often pushed too far, the analogy of treating space as a rubber sheet that can be stretched can be a useful one. Your second question, though, is much more complicated than the first. I am not a physicist, so I won’t be able to fully delve into the math of general relativity and fully answer your question, but maybe I can help a little, if imperfectly.

    Continue to picture space as a rubber sheet. The redshift of distant photons is NOT a Doppler shift. It does not result from the velocity of a receding source. It is the result of the stretching of this rubber sheet of space. The speed of light is a constant. If you stretch the space that a photon is passing through, you also increase the wavelength of the photon. Since c = wavelength * frequency, and c is constant, increasing the wavelength necessarily results in decreasing the frequency, hence the redshift.

    Since the intervening space is expanding, the relationship between the age of an object and its distance is no longer a simple one-to-one relationship. That is, it’s not true that a photon emitted by an object at a distance of 1 billion light years must have been emitted 1 billion years ago. The math is more complex, and unfortunately is beyond my ability to explain. I will defer to others if they want to get into it with you.

  22. #22 Wow
    April 14, 2015

    what conclusions can we draw about their relative positions when we really don’t know where they are now or even if they exist now?

    The same conclusions we can draw from skeet shooting, where you have can’t see where the target is when the bullets hit, so you have to lead the target.

    No different: we do the same work for the slower bullets that is done for light.

    And this is a THIRD type of question, all “supposedly” your same original “problem”.

    So I will conclude you’re not an honest broker and will ignore any questions you give in the future since answering is a waste of time.

  23. #23 A van Wyk
    April 14, 2015

    I tend to favor the “big rip” as the most likely of the fates of our universe, where everything just keeps on accelerating until it reaches a speed faster than the speed of light. 13242033

  24. #24 Michael Kelsey
    SLAC National Accelerator Laboratory
    April 14, 2015

    @Sal Minella #20: You asked, “If we are comparing the redshift of two objects, what conclusions can we draw about their relative positions when we really don’t know where they are now or even if they exist now?”

    If the _ONLY_ data you have redshift, then you cannot directly draw any conclusions at all about positions. If you have other data, then there’s a lot you can derive. So what do we do?

    1) If you have observations of the structure of whatever you’re measuring, then you can _see_ whether it is a single star (within our galaxy), a globular cluster (near our galaxy), or another galaxy out at cosmological distances.

    2) If you can tell that it’s a galaxy, then you can observe other stars in that galaxy, such as Cepheid variables, and use those observations to derive a distance, without any reference to redshift whatever.

    (2) is exactly what Hubble did back in the 1920’s, using Leavitt’s data on Cepheids. He was able to plot two _independent_ observations: distance derived from Cepheid luminosity/period vs. average redshift for each galaxy. The result was an approximately linear relationship with some clustered exceptions.

    Science is NOT a bunch of independent factoids which you can pick and choose from. It is an intricate, tightly woven network of interrelated and mutually consistent observations, inferences, and laws.

  25. #25 Sal Minella
    New York
    April 14, 2015

    Thanks to all of you who have addressed my posts.

    With your indulgence; Information about objects at vastly different distances from Earth is presented as though the objects are being observed in the same time context. From our perspective, we are seeing them at the same time (Earth time), however what we are seeing occurred at vastly different times.

    1) Why isn’t time a consideration in these presentations?
    2) How can comparative conclusions between objects that have considerably time-contexts have any meaning?

  26. #26 Sal Minella
    April 14, 2015

    typo – should read

    2) How can comparative conclusions between objects that have considerably DIFFERENT time-contexts have any meaning?

  27. #27 Michael Kelsey
    SLAC National Accelerator Laboratory
    April 14, 2015

    @Sal #25: You really, *REALLY* should take a decent introductory astronomy course (unless you’re afraid it will affect whatever biases you already have).

    Time is a central feature of any astronomical observation. Thinking it isn’t only reveals your bias, or your ignorance, or your trolling.

    The relationship between time and distance, in astronomical context, is precisely what premits us to infer developomental and evolutionary (ooooh, scary word that!) progression. How do galaxies develop over time? We can learn that by making observations of galaxies at varying distances, and comparing their structure (morphology) and composition, and finding relationships between features and age.

    The physical laws of the Universe, in particular electromagnetism and gravity, have not changed over the course of the ~10 billion years for which we have direct observational evidence. That is not a blind assertion nor a statement of faith, it is a very quick summary of a vast amount of observational DATA. If you want to know why it is true, then it is YOUR responsibility to go read about the science, perhaps take some introductory astrophysics and cosmology courses at your local community college, or just read some of the scientific review papers on the subject.

  28. #28 Sal Minella
    New York
    April 14, 2015

    Michael Kelsey,

    Thanks for your response and your suggestion but not for the time you wasted accusing me of being a troll. In addition you did not answer question 2. Here is what I am getting at: the statement is made often that “objects farther away from us are moving away from us faster than objects that are closer to us ergo the universe is expanding faster as time goes on”. In a time context (as opposed to a distance context), as we observe objects farther back in time (farther away) they are moving away from us faster than objects closer to us in time (nearer to us). It would seem obvious to conclude that objects are decelerating as time goes on, not accelerating yet, that is not the accepted theory. I am not knowledgeable enough to draw the non-obvious conclusion. I appeal your knowledge to explain this seeming discrepancy and to articulate it in a way that a science-curious person could understand. If you aren’t able or just don’t want to take the time, I understand.

  29. #29 Ragtag Media
    April 14, 2015

    @Michael 27
    What about what about the speed of light slowing? I read a number of years about it slowing down, is that true or has it been debunked?

  30. #30 Wow
    April 14, 2015

    Almost nothing evident to show slowing light. It’s usually used by YECers to insist how the earth must be 6000 years old but have a set of stars that are millions of light years away.

    VERY early there’s scope for a small change in the speed of light, but not much, and very uncertain.

  31. #31 eric
    April 14, 2015

    @28: AIUI, if the expansion was decelerating, things at the edge of the observable universe would be coming into our observable universe instead of going out of it.

  32. #32 Sal Minella
    New York
    April 15, 2015


    That’s just exactly my point, the WERE going out of the edge of our observable universe but, what are they doing now?

  33. #33 Wow
    April 15, 2015

    Your problem is you’re using “now” in two different ways at the same time, then wondering why it seems silly.

    The silly isn’t the science, it’s your language use.

    Read the science, stop asking others to do the work for you.

  34. #34 eric
    April 15, 2015

    That’s just exactly my point, the WERE going out of the edge of our observable universe but, what are they doing now?

    I thought your point was ‘how can I tell that the expansion is accelerating rather than decelerating over time?’ That was the point I and several others have answered. What those ‘disappearing’ objects are doing now is irrelevant to your question about the evidence for accelerating expansion.

  35. #35 Ron Pearson
    Bath UK BA1 7TZ
    October 16, 2015

    Dark energy cannot be a proportion of the total since it is totally different from the energy we measure!
    It is useless to speculate on dark energy when the energy creation described by inflation cannot be switched off. This dwarfs any effect of the anti-gravity ascribed to dark energy. See showing how inflation and dark energy have a simultaneous solution -peer-reviewed & published. Dark energy is shown to be sub-quantum by ‘opposed energy dynamics’ that causes each component to feed off the other or mutually annihilate forming a filamentous structure in a state of ever-accelerating expansion. It has self-organising property with filament vibration producing information as quantum waves. In this way dark energy generates everything else. The theory provides new opportunities for theoretical physicists and provides ideas for new experiments in orbit

  36. #36 Wow
    October 16, 2015

    “Dark energy cannot be a proportion of the total since it is totally different from the energy we measure!”

    Is it? How do you know?

    I’d rather not click on some unknown link on the internet. Especially if I don’t know what I’m supposed to find there, crank or genius.

  37. #37 john debaggis
    wrentham mass
    November 21, 2015

    I wanted to post this last week however I wanted to safeguard this idea. anyway this may explain dark energy. think of the universe as a magnet. if mass attracts makes sense empty space repels. so gravity as we know it is only half the story. if you run a model of a galaxy with the gravity we know it has it falls apart. only when we add gravity it works. take that extra gravity apply it from the outside pushing down on matter, a repulsion of empty space. this could explain the extra gravity holding the galaxies together while also explaining why they are being pushed apart. so the gravity we feel is only partly from the attraction of mass the other source is empty space as an external force. so mass is to attraction as empty space is to repulsion equals gravity. ty for reading john debaggis wrentham mass.

  38. #38 Wow
    November 22, 2015

    “take that extra gravity apply it from the outside pushing down on matter,”

    To push requires something to push FROM.

    What is this pushing bracing itself against?

    “so mass is to attraction as empty space is to repulsion equals gravity”

    This would make the amount of space between two objects proportional to the distance between them. Not INVERSELY proportional to the distance between them, as we observe in reality.

    It would ALSO mean there would be no dependency on mass as we also observe in reality.

    For two counts of contra-indicated by evidence, your hypothesis fails.

  39. #39 Wow
    November 22, 2015

    This would make the amount of force between two objects proportional to the distance between them.


New comments have been disabled.