Dark Energy is Real. Really?
I don’t have much time to post today after spending all morning in a meeting about Assuring a Quality Experience in the Graduate College and in between reading project reports this afternoon.
However, I couldn’t resist a quickie just to draw your attention to a cosmology story that’s made it into the mass media, e.g. BBC Science. This concerns the recent publication of a couple of papers from the WiggleZ Dark Energy Survey which has used the Anglo-Australian Telescope. You can read a nice description of what WiggleZ (pronounced “Wiggle-Zee”) is all about here, but in essence it involves making two different sorts of measurements of how galaxies cluster in order to constrain the Universe’s geometry and dynamics. The first method is the “wiggle” bit, in that it depends on the imprint of baryon acoustic oscillations in the power-spectrum of galaxy clustering. The other involves analysing the peculiar motions of the galaxies by measuring the distortion of the clustering pattern introduced seen in redshift space; redshifts are usually denoted z in cosmology so that accounts for the “zee”.
The paper describing the results from the former method can be found here, while the second technique is described there.
This survey has been a major effort by an extensive team of astronomers: it has involved spectroscopic measurements of almost a quarter of a million galaxies, spread over 1000 square degrees on the sky, and has taken almost five years to complete. The results are consistent with the standard ΛCDM cosmological model, and in particular with the existence of the dark energy that this model implies, but which we don’t have a theoretical explanation for.
This is all excellent stuff and it obviously lends further observational support to the standard model. However, I’m not sure I agree with the headline of press release put out by the WiggleZ team Dark Energy is Real. I certainly agree that dark energy is a plausible explanation for a host of relevant observations, but do we really know for sure that it is “real”? Can we really be sure that there is no other explanation? Wiggle Z has certainly produced evidence that’s sufficient to rule out some alternative models, but that’s not the same as proof. I worry when scientists speak like this, with what sounds like certainty, about things that are far from proven. Just because nobody has thought of an alternative explanation doesn’t mean that none exists.
The problem is that a press release entitled “dark energy is real” is much more likely to be picked up by a newspaper radio or TV editor than one that says “dark energy remains best explanation”….
In the Dark 
May 20, 2011 at 6:29 pm
You don’t say “Wiggle-Zed”?
May 20, 2011 at 11:08 pm
I do. But people tell me off.
May 21, 2011 at 9:29 am
I do, it annoys people. I also say “Zed-Zed Top”, that winds then up even more.
May 20, 2011 at 8:25 pm
I thought their main claim was that now we have supernovae and galaxy evidence for the phenomenon. Independent evidence is a good thing. Does the CMBR support Dark Energy?
About all i’ve heard that’s known for sure is that if Dark Energy is real, then it’s repulsive.
May 21, 2011 at 3:29 am
Hi Peter,
Keen reader of your blog of course, so I couldn’t resist posting a reply! Thanks very much for your description of our work, and I certainly 100% agree with your conclusion. The starting point was of course that the cosmological constant currently favoured by the data is somehow a “real material” (whatever that might be) on one side of Einstein’s equations, whereas a modification to gravity would appear on the other side. But that interpretation is certainly open to further debate and exploration by many other datasets and theories.
Media work is something that I find quite difficult personally but nonetheless very interesting. How do we strike the balance between on the one hand wanting high scientific accuracy, which I suspect often requires a lot of background to be put across, but on the other hand boiling things down to a quick, snappy message? The caveats and qualifications required by science often don’t seem to lend themselves well to a “sound bite”.
On the other hand, hopefully even sound bites can put across the idea to the public that in astronomy and cosmology we can try and address some interesting questions which might be worth funding. Although I sometimes don’t know if the preponderance of “dark matter” and “dark energy” in the public domain of cosmology stories appears as a positive thing, in that our data contain real, exciting puzzles to solve about the Universe, or a negative thing, in that it must sometimes appear that we are “making things up” !
Thanks again for your blogs over the years, always interesting.
Chris
May 21, 2011 at 5:29 am
Peter, if scientists didn’t oversimplify they would never get anything in the media. I don’t think I’ve ever seen a scientific press release which a specialist wouldn’t be able to pick holes in.
May 21, 2011 at 9:35 am
Karl,
To simplify is good, but to oversimplify is not. It’s the sort of thing that leads to public distrust of scientists. It’s a fine line, of course, but it’s an important one.
Peter
May 21, 2011 at 5:07 pm
Peter: it’s neither wiggle-zee nor wiggle-zed. You just say wiggles, but possibly with a french accent (or so I thought – surprised Chris Blake didn’t point it out).
May 21, 2011 at 6:29 pm
I could have had a laugh by saying it’s actually pronounced “fanshawe” ..
May 22, 2011 at 11:23 am
yes, its pronounced “wiggles” and this still makes me laugh, for no apparent reason.
May 22, 2011 at 11:32 am
I’m sure I’ve heard it pronounced Wiggle-Zee on more than one occasion. I must have been experiencing auditory hallucinations.
May 22, 2011 at 9:46 pm
Is this the offending article?
http://www.bbc.co.uk/news/science-environment-13462926
May 22, 2011 at 9:51 pm
Ah I see you have already linked, sorry.
I agree entirely with your post however… I am decidedly a layman, but I am irritated by the manner in which dark matter and energy are referred to as fact in the media.
May 22, 2011 at 11:56 pm
Now that we’ve established that it annoys Australians if it’s pronounced Wiggle-Zee or Wiggle-Zed then there’s no question how it should be spoken. 😉
May 24, 2011 at 3:09 pm
I attempted to read the WiggleZ report of the growth of structure, but as an innocent bystander I could get almost nothing from it.
That gravitation would in time increase the relative localization of massive objects while the expansion of spacetime increases the intervening space between material structures composed of clustered galaxies does not surprise me at all, since the localized effect of gravitation diminishes with distance and the effects of expansion on spacetime accumulates, regardless of whether that expansion is accelerating or decelerating.
So, let me go back to the original type Ia SuperNovae studies that concluded that the expansion of the universe is accelerating. In simple terms, standard cosmological models that predicted galaxy distance from redshift agreed with the more reliable estimation of galactic distance based on type Ia SNe luminosity for more recent light emissions from nearby galaxies without using a cosmological constant parameter. However, more distant galaxies’ distance estimates disagreed unless a cosmological constant was used to indicate acceleration.
If I understood correctly, it was the more ancient light emissions from more distant objects that indicated an increased rate of expansion. However, those more ancient light emissions from more distant, high-z objects represent the prevailing conditions looking back to the EARLIER universe, whereas the more recent light emissions from nearer, lower-z objects represent only more recent conditions of expansion.
From that perspective of the observational data, I can only conclude the type Ia SNe data indicated that the expansion of spacetime has DECELERATED, as originally expected, not requiring any dark energy.
Surely I’ve simply misinterpreted something fundamental, but if so no one has yet been courteous enough to explain it to me. Anyone?
May 24, 2011 at 6:18 pm
Thank you very much for your explanation of cosmological models. Believe me, I’m not trying to be obstinate, but it is the paragraph that you didn’t understand and are certain is wrong that is the source of my conflict. The rest was just prolog. Please allow me to try to explain differently.
Astronomers tend to consider they are observing objects, but in fact they are indirectly interpreting the properties of detected light to derive information about the emitting object. For example, the cosmological redshift imparted to distant light does not actually indicate the observed object’s recessional velocity relative to the observer but the expansion of spacetime.
As I understand, as a packet of light traverses expanding spacetime the distance that it must traverse to arrive at any eventual destination (point of detection) is increased and its wavelength is linearly extended – an effect that accumulates in time for each packet of light or detected photon.
A packet of light emitted from a distant object (derived from effects imparted indicating distance traversed) was initially subjected to the rate of spacetime expansion in effect at the moment of emission. For example, a detected photon that was emitted from a galaxy that is 5Glya indicates the prevailing effects of expansion as it traversed expanding spacetime for the past 5 billion years.
A detected photon emitted from a galaxy that is 10Gla indicates not only the same effects of expansion as it traversed expanding spacetime for the past 5 billion years, just like the previous example, but it also reflects the effects of spacetime expansion that prevailed during the PRIOR 10 billion years.
The difference between the redshifts, for example, of the two samples of light is that the light emitted from the more distant galaxy indicates the effects of expansion imparted from 5-10 billion years ago in addition to those imparted for the past 5 billion years.
Since it was the more distant, higher-z observations that indicated a greater rate of expansion – requiring that the Omega and lambda model parameters be used to indicate increased expansion relative to the standard model parameters, it was the rate of expansion that occurred between 5 billion and 10 billion years ago that was greater than the prevailing rate during the past 5 billion years. If this is correct, the data indicates that expansion has decelerated.
Sorry to be so tedious in my explanation, but I can’t even guess where the misunderstanding lies. I really appreciate any additional help you can provide.
May 25, 2011 at 12:51 am
First, I agree this is probably too complex to clarify on a blog & will desist – thanks very much for your responses. I will read your papers with interest.
Just to clarify a bit, as I recall the High-z Supernova Search Team seemed to use the term high-z synonymously with the term distant – that’s how I was using it and low-z (near). Briefly, I understood that they found that low-z SN Ia luminosity based distances agreed with the distances predicted by cosmological models based on the SN’s host galaxy redshift without adding ‘acceleration’. The high-z SN Ia luminosity based distances did not agree with cosmological models unless they imposed acceleration in the models.
That infers to me that the discrepancy between the more directly determined SN Ia distance and the relation between redshift and estimated distance only arose for the high-z or distant observations. Again, it seems to me that the primary distinction between near and far galaxy observations is that the conditions of the earlier universe effect only the farther galaxies. My primary reference is
http://arxiv.org/abs/astro-ph/9805201v1.
Thanks again for your patience – I know my perspective is highly unorthodox at best, wrong at worst, but sometimes that can be useful. I’ll refer to your research report…
May 26, 2011 at 6:30 am
Success! At least, someone who understands the subject I’m attempting to discuss and can consider from both sides to help me understand. Physicists can be so difficult to communicate with, even when (ostensibly) speaking the same language!
Thanks so much for the analogy of the ship sailing past the horizon – that gave me a much better basis for understanding the cosmological models (since I really can’t do the math)! Also I want to commend you for the earlier explanation of redshift as the ratio of emission-apparent scale factors – new to me and very helpful!
[The SN Ia signal used to determine distance is a brief burst (days) peak period emission luminosity. As I understand, the diminishment of its eventual detected/apparent luminosity must accumulate linearly with distance actually traversed through expanding spacetime. That its brief burst of narrow spectrum light is also redshifted I think is strong evidence that redshift is also the accumulated physical extension of light’s wavelength, imparted by the physical extension of intervening spacetime traversed. At any rate, the correlation between luminosity and the distance the light traversed for the SN Ia samples should be nearly linear.]
For these discussions, since the redshift of the SNe host galaxies’ broader spectrum light can be calibrated to the SNe distances, there should be no question as to whether a galaxy is considered ‘near’ or ‘far’.
I don’t follow the meaning of: “In this sense you are right that nearer objects should show the effect, but “near” in this sense actually means the “high-z” objects, while objects of even higher redshift might show less of a discrepancy.”
With your help (after great difficulty locating the untitled Figures 4 & 5) I was able to better understand how the models work (in very general terms). As I understand, if a data set of redshifts were processed by the model, its distance would be derived using a variety of equations intended to represent the temporally varying factors affecting redshift.
Even the referenced report presumes that universal expansion was decelerating (‘naturally’) until at a point in time several billion years ago (determined by observational analyses) when acceleration began.
How can that be scenario be accurately represented by a model with a constant acceleration/deceleration parameter? Since the distance derived from redshift is a function of both expansion and deceleration (indexed by some inferred proxy for time?), how can an ‘acceleration’ parameter that applies increasingly to observations that are more distant AND more ancient accurately represent temporally varying universal conditions of expansion?
It still seems to me that observational perspective is inverted. Can you possibly clarify further? Thanks again for the excellent help you’ve already given me!
May 26, 2011 at 1:08 pm
Great! Sorry about the linear reference – I knew that…
You’ve helped me a great deal already and provided some very interesting reference material. Thanks so much!
I won’t declare conversion yet, but I have one last loose idea: if the expansion of the universe has begun to accelerate at z~-0.7, perhaps it was not due to some new factor, but that the localization (clustering) of matter and expansion of intervening space finally reached a point that eliminated universal gravitation as an effective long range inhibitor to continued expansion. Maybe the gravitational links finally broke, producing the development of large scale localized structures…
May 26, 2011 at 6:38 pm
Couldn’t the relatively recent development of mass structure on the scale of the ‘cosmic web’ significantly affect universal mass density and expansion?
Aren’t both a positive cosmological constant and a negative deceleration parameter necessary fit the data?
Wasn’t the deceleration parameter intended to represent the effect of expansion atrophy as the energy of initial expansion becomes dispersed?
What effect could negative atrophy represent, other than another nonphysical analytical proxy or ‘fudge factor’ like the cosmological constant?
The few observations of SN Ia near the cusp of the apparent transition from deceleration to acceleration seem to indicate a turbulent period. Please see Figure 1.2:
http://www.arxiv.org/abs/1010.1162
Speaking of analytical proxies…
This is only distantly related, but I am much more certain that galactic dark matter was a misconception: the expected Keplerian rotational curve applies only to a relatively sparse configuration of planets, each in effect independently orbiting a dominating mass (the Sun contains 99.86% of Solar system mass). The vast disperse mass of, especially planar disc, galaxies are locally self-gravitating: their orbital velocities are primarily determined by their interactions with neighboring peer masses rather than some central mass emanating gravitational force.
There have been some successful general efforts to explain the rotational characteristics of spiral galaxies using Newtonian dynamics, such as:
http://www.arxiv.org/abs/1007.3778
and some specific cases such as:
http://www.iopscience.iop.org/0004-637X/679/1/373/
These studies indicate that dark matter is not necessary to hold galaxies together: the gravitational effects of distributed masses is sufficient.
Employing galactic disc objects as microlenses have been used to test for the local presence of a dark matter halo:
http://www.arxiv.org/abs/1103.5056
A study of hundreds of discrete Milky Way (ordinary matter) halo objects, including satellite galaxies, globular clusters, and old stars are used to constrain the mass and distribution of a dark matter halo:
http://www.adsabs.harvard.edu/abs/2005MNRAS.364..433B
More interestingly to me, unlike the (self-gravitating) galactic disc, these more distant discrete objects do generally comply with the Keplerian rotational curve! From that direct evidence I infer that it is the independent orbits of discrete objects around a dominating mass that produces orbital velocities diminishing with distance. Distributed mass galaxies should not be required to rotate like sparse planetary systems.