# WMAP Summary: Our Universe

1. Mar 16, 2006

### SpaceTiger

Staff Emeritus
Earlier today, I attended a talk summarizing the latest results from WMAP (three years of data were released) and, I must say, it's very exciting! The standard model has been confirmed to greater accuracy, the problematic aspects of the previous release have mostly vanished, and we have a whole new set of sky maps to look at! I'll summarize what they said at the talk, though I should note here that I have not worked with WMAP, so don't ask me detailed questions. The papers ought to be able to answer those for you.

First of all, the reason it took them so long to process the data was that they were trying to remove the foregrounds in the polarization measurements. The primary foregrounds are from dust and synchrotron radiation, both of which are aligned with the galactic magnetic field. Thus, they had to develop a reliable model both for the magnetic field and the dust.

In fitting to the standard cosmological model, they use six model parameters and they fit to both the temperature and polarization data. The standard model was well fit ($\chi^2\simeq1.04$), as were some of the variations on it. Models using only ordinary (i.e. baryonic) matter were a very poor fit ($\chi^2\simeq250$), and models without reionization and/or inflation were a poor fit as well.

Since the error bars on the matter, baryon, and dark energy content of our universe were small in the first year results, we didn't expect anything groundbreaking in that arena. We appear to live in a universe composed of

~73% dark energy
~23% dark matter
~ 4% ordinary matter

The various surveys disagree at the few percent level about the first two numbers. WMAP alone gives error bars of a few percent on the measurement of the matter content of the universe. The universe is also consistent with being perfectly flat. One way to give this result is to say that the effective energy density of curvature is less about 2%. The best fit model is slightly curved, but this result is not statistically significant.

But that was all old news. What's new? Well, for starters, the total optical depth to the surface of last scattering (where the CMB was created) has been reduced significantly since the first release. This means that they're estimating the first stars to have formed around z~11 (rather than z~17) and we expect that the universe to had begun reionizing itself at around the same time. This is good because it's more consistent with observations of quasars that show partial reionization to be occurring at z~6. This is a result that came from the polarization data, so we wouldn't have had it without the extra time they took to make the release. Other big news is that the primordial power spectrum is more clearly not consistent with scale-invariance. This means, basically, that we have confirmed another prediction of inflation.

In addition to these new measurements, the third-year release features a glitchless power spectrum, less deviation at the low-l multipoles, extremely gaussian fluctuations (another prediction of inflation), model consistency with other CMB experiments, a dark energy equation of state still consistent with -1 (cosmological constant), and even some interesting results about spinning dust grains.

There is much more I could talk about, but I'd rather hear what people think first.

Last edited by a moderator: Mar 26, 2006
2. Mar 17, 2006

### Garth

Thank you ST for that immediate and informed summary.

I notice from the new power spectrum that the quadrupole is still missing, and it looks a bit wobbly at the high-l end. I also understand the 'axis of evil' is still there.

The baryon density looks a little low at 1.5%, is this consistent with BBN deuterium and helium abundances?

I was a little mystified by your statement
I may be ignorant but I thought scale invariance was a prediction of inflation?

Garth

3. Mar 17, 2006

### SpaceTiger

Staff Emeritus
There was no way for the third-year results to remove those anomalies entirely, since the primary limitation in that regime is cosmic variance (i.e. the fact there's only one universe to observe). The anomalies are less deviant than in the previous release, however, and it makes the already a posteriori statistics concerning the "axis of evil" less convincing.

The only big inconsistency is with lithium, which has always been off from predictions. See the paper for more details.

Inflation predicts a nearly, but not exactly scale-invariant spectrum. The deviations expected are at about the level we can now observe.

Last edited: Mar 17, 2006
4. Mar 17, 2006

### Garth

As you know I would prefer the statement:
to read "The universe is also consistent with being perfectly conformally flat."

As the interpretation of the WMAP data, no matter how precise, is model dependent, it would be prudent not exclude other possibilities unnecessarily.

Garth

5. Mar 17, 2006

### PhysicsDilettante

Thank you very much for this summary, ST (your posts are always among the most clear and informative on this site).

A couple of quick clarification questions. What are the "units" being used here so that these add to 100% of the universe's composition? I'm particularly not clear on how dark energy is measured so that it can be added to ordinary and dark matter.

In a similar vein, is there an accounting for energy related to "ordinary" matter? I'm thinking here of, for example, all the photons currently travelling all over the universe. Are they accounted for in the ordinary mass component via E=mc^2?

Thanks.

Last edited by a moderator: Mar 26, 2006
6. Mar 17, 2006

### Garth

Hi PhysicsDilettante!
And welcome to these Forums, (I noticed nobody has done so before me.)
Seconded.
These are straight proportions by mass of the total, measured as densities.

As the universe is flat, it is assumed the total density is that of the critical GR density for closure:
$$\rho_{total} = \rho_{critical} = \frac{3H_0^2}{8 \pi G}$$

with the present value of h = 0.71 this 'closure density' is $9.5 \times 10^{-28}$ gms.cm-3, which is taken as the present density of the universe.
The largest resevoir of photon energy is in the CMB, which is about 2 OOM less than the baryonic density and yes, this is taken into account in that cosmological inventory. Although for most practical problems it is ignored except in the early universe where it becomes dominant.

$$\rho_{matter} \varpropto R^{-3}$$ and $$\rho_{radiation} \varpropto R^{-4}$$

I hope this helps.

Garth

Last edited: Mar 17, 2006
7. Mar 18, 2006

### RandallB

What is a WMAP “axis of evil”? (Nothing to do with Reagan’s use I expect)

Is there a simple definition for, or a link that describes, “Axis of evil”?
What I’ve found so far, assumes I know what it is.

My best guess is it has something to do with after adjusting for all measurably known gravitation movements the CBMR shows up a bit blue toward one direction and redder in the opposite direction to create a bi-pole or “Axis”. That happens to be in alignment with the previously identified Great Attractor in Virgo with Pisces being the opposite pole.

After adjustments which CMBR background axis end, Pisces or Virgo, are they saying appears shifted red and which end blue?

Confusing me is they seem to say Virgo (with the “Attractor”) has the greater redshift. Which to me would mean more and growing separation there than in the blue area, not greater attraction.

8. Mar 18, 2006

### Garth

The 'Axis of Evil' is the apparent non-statistical alignment of the low-l modes with local geometry, possibily with the passage of the galaxy relative to the CMB. It was named after President Bush's famous phrase and the title has stuck.

The extreme view is that this shows that all the CMB anisotropies are local effects and nothing to do with cosmology.

A more reasonable view is that it signals a non trivial topology, or that these low-l modes are actually local contamination.

There is a powerful dipole, about 100 times larger than the CMB anisotropies due to the Earth's motion w.r.t. the Surface of Last Scattering and a local mass, such as a Local Group halo, or larger mass further away, which the galaxy is moving w.r.t., could be lensing this dipole and producing the local aligned low-l mode signals. Local Pancake Defeats Axis of Evil

If this is the case then the power spectrum will really be deficient at this low-l mode end.

Garth

Last edited: Mar 18, 2006
9. Mar 18, 2006

### SpaceTiger

Staff Emeritus
And the most reasonable view (or, rather, most likely interpretation) is that it's an illusion generated by a posteriori statistics. As the WMAP team said in their paper, the effect by itself isn't significant enough to provide evidence for anything. It will have to be coupled with other lines of evidence in order to be meaningful.

10. Mar 19, 2006

### Chronos

The latest candy:

CMB multipole measurements in the presence of foregrounds
http://www.arxiv.org/abs/astro-ph/0603369
Authors: Angelica de Oliveira-Costa (MIT), Max Tegmark (MIT)
. . . Applying our method to the WMAP quadrupole and octopole, we find that their previously reported "axis of evil" alignment appears to be rather robust to Galactic cut and foreground contamination

11. Mar 19, 2006

### Garth

http://lambda.gsfc.nasa.gov/product/map/dr2/pub_papers/threeyear/parameters/wmap_3yr_param.pdf
Page 67
"we could quote the probability of this occurring randomly as less than 2%."
It seems to me that this is actually saying the probalbity of the evidence for non-Gaussiaity, i.e. that the Axis of Evil does exist, is at the 98% level.

However, because that would require the team to rethink all their theories and a "dramatic reinterpretation of our theories of primordial fluctuations" then even more compelling evidence is required.

98% is normally good enough for me!

Garth

Last edited by a moderator: Apr 22, 2017
12. Mar 19, 2006

### marcus

an article by Chris Vale.
thanks for the link, I thought it was quite interesting
also I see that Vale and collaborators have a series of earlier papers about weak lensing by local structure, including some numerical simulations

Vale was at the UC Berkeley physics department but now seems to be at Batavia (Fermilab). Might be interesting to contact him and see if his ideas have progressed as to a possible cause of the "A. of E." alignment.

Last edited: Mar 19, 2006
13. Mar 19, 2006

### RandallB

Maybe one of you guys can help me understand. Before I can even begin to comprehend what is thought to be seen in quadropole and octopole observations, I’m having trouble getting clear what is being said about the dipole.

I understand that taking the CMB as a reference to measure from there is no reason to expect Earth to make uniform measurements of it as though we are in a common “reference frame” with it. This gives us the largest part of the observed CMB temperature variations. At least four things should move us differently than the CMB.
(1) Earth as it moves around the Sun.
(2) The additional motion we have following the Sun’s orbit in the Milky Way.
(3)The motion we get from the Milky Way as a whole being effected by the center of gravity of the galaxies making up our “Local Group” either an ‘orbit’ or ‘drift attraction’.
(4)And finally how our Local Group might move in relation to mass density located around it.

Allow me to define two dipoles as “The dipole anisotropy” and “A CMB dipole”.
Adding up all four elements above gives us “The dipole anisotropy” that they work so hard at removing from the WMAP data. Once removed the background is much more uniform.
BUT I’m unclear; is there still “A CMB dipole” remaining?

This is the part seem to find a clear statement on. Is the dipole considered only an effect of our intrinsic motion, or do at least some claim a “CMB dipole” in the background itself. And if so what is the alignment of that “CMB dipole”, and how close is it to the alignment of “The dipole anisotropy”. Since these should be two very different things, it would seem wrong for them to be in common alignment.
(The center octopole chart in the Chris Vale doc. Seems to show a dipole not in alignment with “The dipole anisotropy”.)

Also, since the adjustments for the four different motions are so important does anyone define their relative speeds and what how each of those speed vectors are aligned on the WMAP galactic chart?

Finally, Since the data must be constantly adjusted for our orbit around the sun there is no fixed vector for that part of “The dipole anisotropy”. But the yearly moving vector should obit in a single plane. Shouldn’t that plane be defined on the WMAP galactic chart?
If only to confirm the plane of the quadropole and octopole measurements are not in alignment with our orbit, since if they were some double checking to confirm those orbit adjustments were not contributing to those readings somehow.

14. Mar 19, 2006

### Garth

Correct
No, not if it is done correctly. But if there is lensing from a 'local' mass then as Chris Vale says:
That is not a dipole in that chart!
The observed dipole is purely the effect of the Earth moving relative to the Surface of Last Scattering (SLS) (The CMB frame)
Earth's velocity around the Sun ~ 10-4c, Sun's velocity around the galaxy ~ 10-3c, Galaxy's velocity wrt SLS ~ 10-3c.

Garth

Last edited by a moderator: Mar 26, 2006
15. Mar 19, 2006

### SpaceTiger

Staff Emeritus
There seems to be some misunderstanding about calculating statistics a posteriori versus a priori. If I'm driving along the highway one day and see a license plate with exactly the same set of letters and numbers as my computer password, I might think, "Wow, what are the chances of my seeing that license plate?" Suppose I actually go home and calculate that probability and find that the chances were 1/1000 (or probably something smaller). Does that mean there was a conspiracy or divine intervention involved? After all, there was only a 1 in 1000 probability of seeing that license plate by chance!

If it's not already obvious, no, this doesn't mean something weird is going on. There's an inherent bias in this hypothetical "experiment" that comes from the fact that you wouldn't have calculated the probability of the event if you hadn't noticed it as being unusual. The same is true for the axis of evil. The people who found it were originally looking for alignments with the galactic and supergalactic planes. They found no such alignment, so they started comparing the low multipoles to other notable celestial structures, including the ecliptic plane. When they found one that matched, they quickly announced it, along with the probability that such an alignment would occur.

The probability we're really interested in, however, is the one that tells us how likely they were to find some alignment, any alignment. Unfortunately, this is virtually impossible to compute and depends more on sociology than statistics or astronomy. This is why, in the world of a posteriori statistics, we usually don't acknowledge something as evidence until it's shown to be extremely improbable (like 99.99%).

Side Note: This tactic is similar to that used by Arp and company when, for example, claiming quasar-galaxy alignments. The "axis of evil" folks are not, by most definitions, cranks or extreme fringe scientists like Arp, but some of their arguments work on a similar principle. It's certainly possible that there is some real contamination in the low multipoles, but the point here is that this evidence isn't strong enough to make dramatic claims.

Last edited: Mar 19, 2006
16. Mar 20, 2006

### Chronos

It's a big universe and chance alignments are to be expected. I agree with ST on that issue. Despite our best efforts, I fully expect 'axis of evil' results. Find them all over the sky, and I will be persuaded to abandon that stance.

17. Mar 20, 2006

### SpaceTiger

Staff Emeritus
Last edited by a moderator: Apr 22, 2017
18. Mar 20, 2006

### Garth

A good example ST, and a historical example was D-Day crosswords are still a few clues short of a solution

He was taken apart by the security services and D-Day was almost cancelled but it was all just coincidence...

The context is crucial.

If I see my Bank card PIN number on the registration plate of the car in front of me I put it down to coincidence, if I see it written down in a colleague's note book I might be more suspicious.

The question that is put to be answered by the statistical analysis is also crucial.

In the 1980's we ran into a lot of trouble here in the UK over the BSE crisis. Lots of cows were falling sick with 'mad cow disease'. The question was could we still eat beef from our herds. The government asked their scientists the wrong question. They asked, "Is there evidence that the meat is unsafe to eat?" And the answer was no, there was (at the time) no evidence that humans could be infected.

However given that cows were literally dropping down on the way to the abattoir, the question should have been: "Is there evidence that the meat continues to be safe for human consumption?" And the answer would have been, "No there is no evidence that the meat is still safe." They didn't ask this question because of the vested interest.

The sentence on page 68 of http://lambda.gsfc.nasa.gov/product/map/dr2/pub_papers/threeyear/parameters/wmap_3yr_param.pdf
smells of vested interest to me..

Garth

Last edited by a moderator: Apr 22, 2017
19. Mar 20, 2006

### SpaceTiger

Staff Emeritus
Good, then we're agreed that these results don't constitute "98% evidence" that they need to rethink their theories, as you said.

But that's exactly the point -- it depends on things that cannot be easily computed. Giving statistics for the "axis of evil" is extremely deceptive.

So you're saying it's not in their interest to report a ground-breaking result? I find that very odd. Smells of a conspiracy theory to me.

20. Mar 20, 2006

### RandallB

Does anyone know of a site with a diagram of standard sky chart reference points plotted onto a Galactic Chart?
Mainly the Ecliptic; Earths equator, North, & South Poles; and a few of the key constellations or reference stars.
When something is described as near the ecliptic on a galactic chart, I can’t find anything to compare with to know where that ecliptic is.

Also does the galactic chart have a standard coordinate location convention similar to right ascension and declination?

Last edited: Mar 20, 2006
21. Mar 20, 2006

### Garth

Touche!

If you ask the question:"Are the Three Year WMAP results consistent with the mainstream $\Lambda$CDM model?" The answer is Yes.

If you ask the question:"Is there evidence that the low-l mode anisotropies in that data are aligned?" The answer is yes, "the probability of this occurring randomly is less than 2%".

It depends on the question you ask.

Garth

Last edited: Mar 20, 2006
22. Mar 21, 2006

### Chronos

Would it be fair to suggest this might be a selection effect, or accumulation of systematic errors? Given the otherwise robust results favoring the concordance model, I am reluctant to disregard these alternatives.

23. Mar 21, 2006

### Garth

Are these two answers consistent with each other?

The effect of admitting the AOE exists is that is enhances the deficiency of the low-l mode anisotropies and that would be evidence for non-Gaussality. From that 3yrd year WMAP 'Cosmological Implications' paper:
(Emphasis mine)

Already the quadrupole is deficient, and these further deficiencies may be the evidence that Komatsu/Colley & Gott did not find.

As I have said the data is not only consistent with a flat universe but also a conformally flat one, a model not yet considered as it entails modification of the GR/Friedmann models.

Garth

Last edited: Mar 21, 2006
24. Mar 23, 2006

### Garth

From the OP:
This seems to be rather a low baryonic density, especially as it may be the case that more than this has already been observed as WHIM.

The visible mass density is only $\Omega_v = 0.003$, whereas $\Omega_b$ is about an OOM greater than this, so where are the 'missing baryons'?

In the paper that proposed these missing baryons were to be found in the warm-hot ionized medium (WHIM),http://lanl.arxiv.org/abs/astro-ph/0501126", Nicastro, Elvis, Fiore & Mathur seem to have found too much!

As I have posted before in #15 on the Self Creation Cosmology thread, their paper states:
Now to interpret this result [O/H] is needed; and in Table 1 they state at:
z = 0.011 [O/H] > -1.47 and at
z = 0.027 [O/H] > -1.32,
so the upper limit is:
$\Omega_b$WHIM > 4.3 × 100.47 % = 12.6%
and the lower limit:
$\Omega_b$WHIM > 1.3 × 100.32 % = 2.7%.

Which was indeed consistent with the old standard model BBN of about $\Omega_b = 0.04$, but even their lower limit seems too high for the new WMAP3 of $\Omega_b = 0.015$.

Garth
[Moderator note: I have edited the OP, to correct ST's error.]

Last edited by a moderator: Apr 22, 2017
25. Mar 23, 2006

### SpaceTiger

Staff Emeritus
You're right Garth, I just copied the wrong number. The WMAP paper reports:

$$\Omega_bh^2 = 0.0223$$

With h~0.75, this leads to almost exactly 4%.