In Part 1, the measurement of cosmological distances using standard candles was introduced. In Part 2 a couple of types of standard candles, Cepheids and SN Type 1a were described. Part 3 brought that together with a primer on measuring velocity, and how that implies a distance in an expanding universe. With two independent ways of measuring ‘distance’ – Hubble flow and the SNe Ia standard candles, one compares the two and finds that the SN Ia distances are larger than those derived from the Hubble flow – the so called Hubble Residual (HR). Since the later depend on the model of the universe, in this case uniform expansion, the former implies that the Universe is not only expanding, but accelerating. Since we don’t know how that can happen, we invented dark energy to explain it. In this final installment, I’ll look at some challenges to that conclusion; I don’t intend to prove or disprove it – if I could do that, I’d be off claiming my Nobel Prize – but rather maybe illustrate the scientific method – the SCIENCE is never settled.
With that: given the implication/significance of an accelerating universe, there’s been a fair amount of work, even in the discovery papers, trying to find alternative hypotheses.
One possibility, that I’ll sort of dismiss out of hand, not because it’s not necessarily believable but because it has no predictive power, is that physical laws and/or constants are not constants. In that case physics may be different and identical initial conditions generating a Type 1a SNe but at cosmological distances might have systematically different behavior. In addition to the fact that there are no ‘easy’ testable predictions, the SNe used in the discovery of acceleration, while distant, are not that distant cosmologically speaking and therefore we are not looking too far back in time – (remember that owing to the finite speed of light, distance is directly proportional to how far back in time we are looking) and there is no evidence for any change in physics or constants over that “short” period of time. So almost everyone is comfortable assuming that physical laws and constants don’t change in any way that’s important to this problem. This is really a consensus in the true sense of the word.
One other possibility is the affect of dust. Dust, mostly in the form of carbon and various forms of silicate materials is ubiquitous and complicates the interpretation of most astronomical observations. Dust interacts with light by scattering and absorbing it. Therefore, any dust along the line of sight to any particular SNe would have the affect of making it appear dimmer, and so is a natural, systematic (always makes them dimmer, never brighter) bias that, while mundane, could explain for the observations. Unfortunately, there are two major reasons to fairly confidently dismiss this possibility. First, dust absorption and scattering change the frequency distribution of light, the relative amount of energy transmitted at short wavelengths vs long wavelengths.
Think of a sunset – the reason a sunset appears red, pink, orange is that the path through the atmosphere that sun light must traverse is much longer at the horizon than at zenith. The dust in the atmosphere preferentially scatters and absorbs blue light (also why the sky “is” blue), leaving only the reds and oranges to make it to your eye-hole. Astronomical dust has many of the same properties and hence light from the SNe passing through a long path length of dust to get to our telescopes will preferentially have more blue light removed, changing the relative brightness different wavelengths. This particular effect is not observed in the SNe data; the brightness seems to be uniformly suppressed. “Exotic” dust that doesn’t have this wavelength dependent effect (so-called “gray” dust) but only reduces the brightness has been proposed. Realistic astronomical dust models (different composition and shapes) can produce this sort of behavior. But in order to have a 20% affect, very large quantities would be required – far beyond what would have significant effects in other observational data, effects that are not seen. Additionally, one requires a very large quantity of things like iron to be present in this sort of dust, more than is realistically available to make dust. So there’s a lot of detail there, but the upshot is that it is pretty well established at this point that dust is an unrealistic explanation. Probably real consensus here too!
The final thing we’ll look at here, and the subject of the papers that lead to this whole fiasco of a series, is a change in the intrinsic light curve corrected brightness of Type 1a SNe – while we dismissed this affect as following from a change in fundamental physics, that’s not the only way such a change could happen in the distant (old) universe. Type 1a SNe don’t happen in isolation. They occur in a galaxy and their properties are dependent on the stellar population from which they emerge. The ages of stellar populations, their initial composition, the density of the environment from which the form, etc. are all factors that can have subtle and systematic effects on the exact behavior of the eruption. Therefore is not only plausible, but expected, that, if galactic stellar populations (galaxy structure) change with time, Type 1a SNe ‘intrinsic’ brightness will most certainly also evolve. Up to here, nothing controversial – it is well accepted, a scientific consensus even, that this sort of luminosity evolution is not only possible, but actually observed. Indeed, there is an observed correlation between the age of the stellar population and the HR. The original discovery paper (and many follow-ups) looked into luminosity evolution of Type 1a SNe and concluded that is could not be used to explain the observations; there was no evidence for luminosity evolution in the data they used. Follow-up papers have claimed that the age ranges in the sample used to ‘discover’ the acceleration are insufficient to explain the size of the effect. Of course, that conclusion depends on both accurate age dating of the sample as well as accurate calibration of the relationship.
That brings us to the meat of why I started this. There has been a series of papers over the last couple of years that address both these issues. They have been largely from a well respected (as far as I can tell) Korean group. There have been responses and the details have been argued back and forth. I’ll avoid too much detail here as one can very rapidly get underwater. Briefly, the Korean researches initially got very accurate spectroscopy (see Part 3) of a large sample of galaxies that allowed them to very accurately age date them. Then they looked the HR vs age in this sample and found a steeper relationship than previously found. In a series of back-and-forth papers, they claim that the luminosity evolution of SN Ia is significant enough, given the age ranges and the steeper relationship, to explain all of the HR and hence negate the need to acceleration and dark energy. There has been push back of course, including an analysis claiming to show that relation was very shallow and unable to explain the full HR. The responses reviewed the techniques and data sample used in Rose et al 2020 and found that the sample they used was not of high enough quality to get good age estimates of the galaxies in SN Ia samples; when they cross matched with their higher quality data, the found significant over estimates of the ages of the host galaxies thus biasing the aga-HR correlation low. Further, the statistical techniques used have well known biases that were not accounted for. They reanalyzed the exact same data as Rose et al 2020 and confirmed their earlier results. Their results are summarized in this plot: They can easily reproduce the observed Hubble Residual without invoking any esoteric physics like Dark Energy.
Conclusion – is the universe accelerating? Is their such a thing as dark energy? Maybe not. There are much more mundane explanations embedded in the minutiae and details of doing science – the boring stuff – whether or not this particular line of reasoning about luminosity evolution is correct or not. Short of it – accelerating universe is not a ‘fact’ as of yet. There’s a lot of room for the observations that drive it to be wrong, or more precisely the interpretation of those observations to be missing essential context. I can’t say I know the answer, though I suspect the Korean group is correct but that’s partly my contrarian nature. But a lot of science/scientist like to focus on “big” things and can lose sight of the assumptions that underlie the big ideas; no intention necessary – they just become ingrained in our understanding. I’m sometimes amazed how we can sometimes take as established truth things that rely on a single off-hand or unsupported statement from years ago.
Unfortunately, dark energy is sort of accepted as the default baseline paradigm and it’s presented as more of a consensus in the popular press than it really is. It’s hard for practitioners in the field to argue with the cache of a Nobel prize winner, let alone the general public. While scientific consensus is real and valuable if it is true consensus, I personally look at the consensus of dark energy as drifting into consensus of experts or social pressure, at least in the public presentation, but on the professional side as well perhaps. The cynic might observer that there’s lots of funding for experiments to “find the dark energy” – Scientists love them some funding! On the positive side, the scientific process continues. People are looking very carefully at the data and looking at the very basis of the argument for dark energy with a skeptical eye. And with some success. They are publishing in the mainstream journals and are able to present the evidence freely and, even though there is always some pressure to accept the baseline of knowledge (and that’s healthy for well established principles as long as the rare brilliant iconoclast can question and flourish to some degree) in any field, they are not ostracized, censored, don’t have their PhD’s or licenses revoked. Would that this was true in some other fields of scientific endeavor.