Planck 21 March 2013

[Nicolaas Vroom]

Planck 2013 results. XVI. Cosmological parameters document:
http://planck.caltech.edu/pub/2013results/Planck_2013_results_16.pdf
is excellent. It shows the state of art in cosmology at its highest level.
What makes this document so powerfull is that it is easy to go to
the underlying documents (using the links) to get more specific information.

Still I have a question.
At the beginning of paragraph (page 25) I read:
"Unlike CMB data, traditional astrophysical data sets -- e.g., measurements
of the Hubble parameter, type Ia supernovae (SNe Ia), and galaxy redshift
surveys -- involve complex physical systems that are not understood
at a fundamental level."
I agree that the 3 examples are complex physical systems and maybe are
not understood at fundamental level but that does not mean that this is
not true for the CMBR.
What the documents show is that we can measure the CMBR at an extremely
high accuracy and that is a hugh achievement but that does not mean
that ONLY by observing the CMBR we can predict the evolution of the Universe
specic the parameters of the Friedmann equation.
In fact I think that it is easier to calculate the cosmological parameters
by ONLY using type 1A SN data, because observations are available at many
different distances, compared with using ONLY CMBR data because only one
set of data is available for one distance.
For CMBR data you should also compare Planck_only with WMAP_only results.
In paragraph 5.4 Type 1a SN are discussed.
In equation 56 the magnitude parameter is MB is calculated as a function
of H0.
I do not understand this because IMO H0 (Hz z=0) should be calculated
to gether with all the other parameters based on SN observations.
In the document
http://users.telenet.be/nicvroom/friedmann%27s%20equation.htm
I follow that approach.
The same problem I also see in paragraph 5.2 BAO and 5.3 H0
In order to calculate the parameters Omega(L) Omega(m) Omega(k) + Age
based on the power spectrum you should combine the two.
You can by pass the problem by using some iteration scheme and use
H0 as an initial value (I think)

We are living in an interesting world full of surprises.

Nicolaas Vroom
http://users.pandora.be/nicvroom/

[Mod. note: non-ASCII characters removed -- mjh]

[Phillip Helbig---undress to reply]

In article <mt2.0-27426...@hydra.herts.ac.uk>, Nicolaas Vroom

<nicolaa...@pandora.be> writes:

> "Unlike CMB data, traditional astrophysical data sets -- e.g., measurements
> of the Hubble parameter, type Ia supernovae (SNe Ia), and galaxy redshift
> surveys -- involve complex physical systems that are not understood
> at a fundamental level."
> I agree that the 3 examples are complex physical systems and maybe are
> not understood at fundamental level but that does not mean that this is
> not true for the CMBR.

The statement, which I agee with, is that the physics of the CMB is
better understood than that of, say, supernovae.

> What the documents show is that we can measure the CMBR at an extremely
> high accuracy and that is a hugh achievement but that does not mean
> that ONLY by observing the CMBR we can predict the evolution of the Universe
> specic the parameters of the Friedmann equation.

No-one has claimed that.

> In fact I think that it is easier to calculate the cosmological parameters
> by ONLY using type 1A SN data, because observations are available at many
> different distances, compared with using ONLY CMBR data because only one
> set of data is available for one distance.

Within the context of Friedmann-Lemaitre cosmological models, this
doesn't matter. (Of course, in order to calculate lambda and Omega from
the m-z diagram, you need several redshifts, but in theory (i.e. if
measurements had no uncertainties), just one object would be enough to
calculate the Hubble constant.

[Mod. note: non-ASCII character removed. Please post in plain ASCII -- mjh]

[Nicolaas Vroom]

> In article <mt2.0-27426...@hydra.herts.ac.uk>, Nicolaas Vroom
> <nicolaa...@pandora.be> writes:
>
> > I agree that the 3 examples are complex physical systems and maybe are
> > not understood at fundamental level but that does not mean that this is
> > not true for the CMBR.
>
> The statement, which I agee with, is that the physics of the CMB is
> better understood than that of, say, supernovae.

SN Data does not require Power Spectrum nor the program CAMB
See also below.

> > What the documents show is that we can measure the CMBR at an extremely
> > high accuracy and that is a hugh achievement but that does not mean
> > that ONLY by observing the CMBR we can predict the evolution of the Universe
> > specic the parameters of the Friedmann equation.
>
> No-one has claimed that.

I agree.
The issue is the colums 2 and 3 of table 2 (Planck results 16) page 11.
What these columns show are 6 + 20 parameters which apparently can
be calculated ONLY by using Planck data.
Those parameters include H0 and the Age of the Universe.
By Planck data I mean the power spectrum.
I find this quite remarkable (If my understanding is correct) because
this requires that we know (in detail) the relation between all those
parameters and the shape of the power spectrum.
(It is also not simple)

> > In fact I think that it is easier to calculate the cosmological parameters
> > by ONLY using type 1A SN data, because observations are available at many
> > different distances, compared with using ONLY CMBR data because only one
> > set of data is available for one distance.
>
> Within the context of Friedmann-Lemaitre cosmological models, this
> doesn't matter. (Of course, in order to calculate lambda and Omega from
> the m-z diagram, you need several redshifts, but in theory (i.e. if
> measurements had no uncertainties), just one object would be enough to
> calculate the Hubble constant.

Only in theory. If you use the andromeda galaxy you are completely wrong.
My understanding of SN data is that much less parameters are required
to calculate the cosmological parameters, making it more simple.

Nicolaas Vroom
http://users.pandora.be/nicvroom/

[Phillip Helbig---undress to reply]

In article <mt2.0-13764...@hydra.herts.ac.uk>, Nicolaas Vroom

<nicolaa...@pandora.be> writes:

> > > I agree that the 3 examples are complex physical systems and maybe are
> > > not understood at fundamental level but that does not mean that this is
> > > not true for the CMBR.
> >
> > The statement, which I agee with, is that the physics of the CMB is
> > better understood than that of, say, supernovae.
>
> SN Data does not require Power Spectrum nor the program CAMB

True, but that does not mean that they are understood. CAMB is just one
code (though a good one). The basic physics is well understood. Not
everything is known about supernovae.

> The issue is the colums 2 and 3 of table 2 (Planck results 16) page 11.
> What these columns show are 6 + 20 parameters which apparently can
> be calculated ONLY by using Planck data.
> Those parameters include H0 and the Age of the Universe.

One can of course calculate H0 and the age of the universe (which needs
three parameters, e.g. H0, lambda0 and Omega0) without the CMB.

> By Planck data I mean the power spectrum.
> I find this quite remarkable (If my understanding is correct) because
> this requires that we know (in detail) the relation between all those
> parameters and the shape of the power spectrum.
> (It is also not simple)

It is not simple, but it is well understood. There has been a HUGE
amount of research on this in the last 25 years or so. Check out

http://background.uchicago.edu/

and especially

http://background.uchicago.edu/~whu/metaanim.html

> > > In fact I think that it is easier to calculate the cosmological parameters
> > > by ONLY using type 1A SN data, because observations are available at many
> > > different distances, compared with using ONLY CMBR data because only one
> > > set of data is available for one distance.
> >
> > Within the context of Friedmann-Lemaitre cosmological models, this
> > doesn't matter. (Of course, in order to calculate lambda and Omega from
> > the m-z diagram, you need several redshifts, but in theory (i.e. if
> > measurements had no uncertainties), just one object would be enough to
> > calculate the Hubble constant.
>
> Only in theory. If you use the andromeda galaxy you are completely wrong.

ONE object, but not ANY object. Measuring the redshift is easy. You
need an object whose peculiar velocity is small compared to its cosmic
velocity (no problem if its redshift is large enough) and whose distance
is well known (this is the hard part).

> My understanding of SN data is that much less parameters are required
> to calculate the cosmological parameters, making it more simple.

This is true in a literal sense. However, the idea that supernovae are
standard candles is based on observations more than theory. (Actually,
they aren't standard candles, but the idea is that one can calculate the
absolute luminosity based on other observational quantities.) The CMB,
on the other hand, is understood from first principles.

[Steve Willner]

In article <mt2.0-13764...@hydra.herts.ac.uk>,

Nicolaas Vroom <nicolaa...@pandora.be> writes:
> The issue is the colums 2 and 3 of table 2 (Planck results 16) page 11.
> What these columns show are 6 + 20 parameters which apparently can
> be calculated ONLY by using Planck data.

Bad placement of "only." The parameters can be calculated without
use of any data other than Planck temperatures, if that's what you
want to do, but many of the parameters can be measured in other ways,
too.

Columns 2 and 3 show the results for Planck temperature mapping
alone. Combining with _some_ of the available other data gives
columns 4-5 and 6-7. To make the most refined estimate, one would
combine with still other data such as SNe, BAO, and local
determinations of the Hubble constant. I don't see a relevant table,
but some of the figures (e.g., 36, 38) show results.

> My understanding of SN data is that much less parameters are required
> to calculate the cosmological parameters, making it more simple.

It's simpler to understand, for sure, but also gives less
information. Specifically, if properly calibrated -- no easy task!
-- a SN measurement gives the luminosity distance at a given
redshift. That's useful but far from being all of cosmology.

--
Help keep our newsgroup healthy; please don't feed the trolls.
Steve Willner Phone 617-495-7123 swil...@cfa.harvard.edu
Cambridge, MA 02138 USA

[Phillip Helbig---undress to reply]

In article <mt2.0-11669...@hydra.herts.ac.uk>, Steve Willner

<wil...@cfa.harvard.edu> writes:

> In article <mt2.0-13764...@hydra.herts.ac.uk>,
> Nicolaas Vroom <nicolaa...@pandora.be> writes:
> > The issue is the colums 2 and 3 of table 2 (Planck results 16) page 11.
> > What these columns show are 6 + 20 parameters which apparently can
> > be calculated ONLY by using Planck data.
>
> Bad placement of "only." The parameters can be calculated without
> use of any data other than Planck temperatures, if that's what you
> want to do, but many of the parameters can be measured in other ways,
> too.

I realized this a few hours after I had replied. When proof-reading
someone else's text, this is one of my "pet peeves". It is a common
error, even among native speakers of English (perhaps even more common
among them). Putting "only" or "just" in the wrong place changes the
meaning. The fact that he capitalized it, ONLY, re-inforced the wrong
intention. Calculated ONLY by using Planck data: no other way to
calculate it. Calculated by using ONLY Planck data: possible to
calculate it with Planck data and nothing else. The latter could also
be: calculated by using Planck data only.

[Eric Flesch]

On Wed, 27 Mar 13, Phillip Helbig wrote:
> The latter could also be: calculated by using Planck data only.

Yes, I find when writing documentation that putting ONLY at the end of
the sentence is often the only unambiguous placement.

[Mod. note: OK, that's probably about enough about English grammar --
back to the Planck results -- mjh]

[Nicolaas Vroom]

Op maandag 25 maart 2013 22:26:46 UTC+1 schreef Phillip Helbig---undress to reply het volgende:

> In article <mt2.0-13764...@hydra.herts.ac.uk>, Nicolaas Vroom
>
> <nicolaa...@pandora.be> writes:
>
> > SN Data does not require Power Spectrum nor the program CAMB
>
> True, but that does not mean that they are understood. CAMB is just one
> code (though a good one). The basic physics is well understood. Not
> everything is known about supernovae.

The biggest problem is the relation between Magtitude Luminosity and
distance. The problem is that not all light transmitted reaches the observer.
The cause is that light does not always travel in straight lines but
is scattered and bended. IMO the same problem exists for CMBR
and maybe severe)
The issue is not the code of the program CAMB, but the physics behind the code

> > The issue is the colums 2 and 3 of table 2 (Planck results 16) page 11.
> > What these columns show are 6 + 20 parameters which apparently can
> > be calculated ONLY by using Planck data.
> > Those parameters include H0 and the Age of the Universe.
>
> One can of course calculate H0 and the age of the universe (which needs
> three parameters, e.g. H0, lambda0 and Omega0) without the CMB.

That is not the issue.
The issue is that you should have one experiment to calculate the
cosmological constants or parameters. That means only CMBR data or only
SN data.
The problem is that the Hubble parameter is a function of distance or z.
That means you need the friedmann equation to calculate this function.
By inserting z=0 or d=0 in this function you can calculate H0.

> It is not simple, but it is well understood. There has been a HUGE
> amount of research on this in the last 25 years or so. Check out
> http://background.uchicago.edu/
> and especially
> http://background.uchicago.edu/~whu/metaanim.html

http://background.uchicago.edu/~whu/animbut/anim1.html claims:
"Increasing the baryon, or ordinary matter density, modulates the heights
of the peaks"
How do you know that the simulation is correct, ie. describes the physical
reality correct? The basics are physical assumptions (oscillations) but
how do you know if this model correct ?
In fact you need such a model to calculate all 26 parameters and that is
not simple.

> > My understanding of SN data is that much less parameters are required
> > to calculate the cosmological parameters, making it more simple.
>
> This is true in a literal sense. However, the idea that supernovae are
> standard candles is based on observations more than theory. (Actually,
> they aren't standard candles, but the idea is that one can calculate the
> absolute luminosity based on other observational quantities.) The CMB,
> on the other hand, is understood from first principles.

The basic idea is that all SN type 1A are identical, have the same Luminosity
Of course that is not 100% true.
The CMB requires harmonics:
http://background.uchicago.edu/~whu/intermediate/harmonic.html
Also not easy.

Nicolaas Vroom
http://users.pandora.be/nicvroom/

[Phillip Helbig---undress to reply]

In article <mt2.0-30600...@hydra.herts.ac.uk>, Nicolaas Vroom

<nicolaa...@pandora.be> writes:

> > True, but that does not mean that they are understood. CAMB is just one
> > code (though a good one). The basic physics is well understood. Not
> > everything is known about supernovae.
>
> The biggest problem is the relation between Magtitude Luminosity and
> distance. The problem is that not all light transmitted reaches the observer.
> The cause is that light does not always travel in straight lines but
> is scattered and bended.

This is not an issue for the SNIa stuff. Yes, such effects exist, but
the relatively low redshift means that they aren't that important.
Nevertheless, Perlmutter at all examined how their results changed even
in the extreme empty-beam case---not enough to affect their main
conclusions.

> IMO the same problem exists for CMBR
> and maybe severe)
> The issue is not the code of the program CAMB, but the physics behind the code

But lensing is included in CAMB.

> The issue is that you should have one experiment to calculate the
> cosmological constants or parameters. That means only CMBR data or only
> SN data.

Why? I would think the more the better.

> The problem is that the Hubble parameter is a function of distance or z.

What one calculates is H0, which is not a function of distance of z.

> That means you need the friedmann equation to calculate this function.

One needs to know the other parameters to properly interpret the m-z
diagram at higher redshift, but it is not correct to say that the Hubble
constant depends on z. Yes, H calculated at any cosmic time can vary,
and z corresponds to lookback time, but such calculations are set up so
that H0, lambda0 and Omega0 is what comes out.

> How do you know that the simulation is correct, ie. describes the physical
> reality correct?

This question could be addressed at ANY scientific measurement.

[Nicolaas Vroom]

Op woensdag 27 maart 2013 23:05:49 UTC+1 schreef Phillip Helbig---undress to reply het volgende:

> In article <mt2.0-30600...@hydra.herts.ac.uk>, Nicolaas Vroom
> <nicolaa...@pandora.be> writes:
>
> > The biggest problem is the relation between Magtitude Luminosity and
> > distance. The problem is that not all light transmitted reaches the
> > observer.
> > The cause is that light does not always travel in straight lines but
> > is scattered and bended.
>
> This is not an issue for the SNIa stuff. Yes, such effects exist, but
> the relatively low redshift means that they aren't that important.
> Nevertheless, Perlmutter at all examined how their results changed even
> in the extreme empty-beam case---not enough to affect their main
> conclusions.

The issue is to calculate the cosmological constants as described in the
book Relativity by Ray d'Inverno.
See also: http://users.telenet.be/nicvroom/friedmann%27s%20equation.htm
One strategy is to use type 1A SN data.
A different is to use CMBR data.
The question is which one is the best, the most accurate.
See also below.

> > The issue is that you should have one experiment to calculate the
> > cosmological constants or parameters. That means only CMBR data or only
> > SN data.
>
> Why? I would think the more the better.

I agree. In the first phase you should study the parameters based on one startegy only and see (waar het schip strand) where the schip runs ashore.
That means only Planck data, only WMAP data, only SN data, only lensing
data etc.
In a second phases you can combine the results of different strategies.
In the first phase you should calculate the 20 + 6 parameters previous
mentioned. Specific you should not mix strategies.
For example if you study Planck data only you should not include the parameter
H0 calculated based on SN data (or if you do you should mention this).
Table 2 the columns 2 and 3 (Planck results 16) shows only Planck data.
The question is my understanding correct that you can calculate the
parameters H0 and Age from Planck data only.

This question is important because if the parameters H0 and Age come from
a different strategy it is important to know if they are required in order
to calculate the first 6 parameters

SN Type 1A does not allow you to make a difference between baryonic matter
and Cold dark matter (as far as I know). Only the sum of the two is used.
That means that SN Type 1A data does not calculate the parameters:
Omega(b)*h^2 and omega(c)*h^2. (but only uses omega(matter))
This is okay as long as you can calculate the cosmological constants.

Nicolaas Vroom
http://users.pandora.be/nicvroom/

[Phillip Helbig---undress to reply]

In article <mt2.0-22101...@hydra.herts.ac.uk>, Nicolaas Vroom

<nicolaa...@pandora.be> writes:

> > > The cause is that light does not always travel in straight lines but
> > > is scattered and bended.
> >
> > This is not an issue for the SNIa stuff. Yes, such effects exist, but
> > the relatively low redshift means that they aren't that important.
> > Nevertheless, Perlmutter at all examined how their results changed even
> > in the extreme empty-beam case---not enough to affect their main
> > conclusions.

Should be "Perlmutter et al.".

> One strategy is to use type 1A SN data.
> A different is to use CMBR data.
> The question is which one is the best, the most accurate.
> See also below.
>
> > > The issue is that you should have one experiment to calculate the
> > > cosmological constants or parameters. That means only CMBR data or only
> > > SN data.
> >
> > Why? I would think the more the better.
> I agree. In the first phase you should study the parameters based on one startegy only and see (waar het schip strand) where the schip runs ashore.
> That means only Planck data, only WMAP data, only SN data, only lensing
> data etc.
> In a second phases you can combine the results of different strategies.
> In the first phase you should calculate the 20 + 6 parameters previous
> mentioned. Specific you should not mix strategies.
> For example if you study Planck data only you should not include the parameter
> H0 calculated based on SN data (or if you do you should mention this).
> Table 2 the columns 2 and 3 (Planck results 16) shows only Planck data.
> The question is my understanding correct that you can calculate the
> parameters H0 and Age from Planck data only.

OK, I see what you mean now. I agree. When discussing the cosmological
parameters, it is clearer to say "test A says this", and "test B says
this" and "the joint constraints look like this" and so on. Of course,
combining datasets by multiplying likelihoods is meaningful only if the
best-fit values of one test are within the uncertainties of the other
and vice versa.

> This question is important because if the parameters H0 and Age come from
> a different strategy it is important to know if they are required in order
> to calculate the first 6 parameters

One can in principle calculate everything from just the CMB. Of course,
some parameters are constrained better than others.

> SN Type 1A does not allow you to make a difference between baryonic matter
> and Cold dark matter (as far as I know). Only the sum of the two is used.

Right.