A problem with the analysis of type Ia supernovae

[Michael]

 

[RealityCheck01]

An important point from the CosmoQuest discussion of [1711.11237] A problem with the analysis of type Ia supernovae may have been missed.
There is an obvious and fatal problem with the paper that people can easily find by reading the paper carefully. The paper tries to look at the variation of light curves with redshift. There is a template file with an average light curve used in the SALT2 calibration model. The paper fits raw supernova light curves to that average light curve in the template file thus losing the width variation with redshift between the curves! The plot of width versus redshift in the paper is the statistical differences between the raw light curves and the average light curve. As expected we see a random looking scatter.

This is very easy to understand by thinking about a hypothetical of every light curve having exactly the same shape and different widths. An average curve would be that shape with an average width. Fitting any light curve to the average curve would give that average width.

Other problems:
The time dilation of light curves is also measured using non-SALT2 calibration.
The time dilation in type 1a supernova is confirmed by the dilation of their aging, i.e. how their spectra changes during the supernova.
The paper cannot conclude that there is a static universe from 1 line of evidence while ignoring the others (also a hint of the fallacy of false dichotomy).

 

[Michael]

An important point from the CosmoQuest discussion of [1711.11237] A problem with the analysis of type Ia supernovae may have been missed.
There is an obvious and fatal problem with the paper that people can easily find by reading the paper carefully. The paper tries to look at the variation of light curves with redshift. There is a template file with an average light curve used in the SALT2 calibration model. The paper fits raw supernova light curves to that average light curve in the template file thus losing the width variation with redshift between the curves!

Ya, and that's the error in the methodology that David cited.

 

[RealityCheck01]

Ya, and that's the error in the methodology that David cited.

That is good - like I said the flaw is fairly obvious so that paper is just wrong to people who read it.

 

[RealityCheck01]

This whole conversation about Olber's paradox is a great example of why astronomy is still stuck in the dark ages....

That is wrong. Olbers' paradox is a great example of applying simple logic to the universe. It can be stated simply, e.g. in a "static, infinitely old universe with an infinite number of stars distributed in an infinitely large space" every line of sight will end on a star and so the night sky cannot be dark. Or we can do the math and get the same result. The reasonable conclusion is that one or more of the assumptions is wrong. The universe is finite, not static, not eternal or a combination of them.

The "shell of stars" argument is used for the homogeneous universe that we observe at large scales. That allows the argument to be understood by anyone. Add inhomogeneity and obviously some shells will be brighter, some shells will be less bright and we get a non-uniformly bright night sky that still does not exist. But give that inhomogeneity a fractal dimension of 1 and the paradox vanishes! The problem is that observations suggest a fractal dimension of 2.

Olbers' paradox is the earliest evidence that the universe could be expanding or contracting or have an age or be finite. It is minor. The There is overwhelming other empirical evidence that shows that the universe expanded from a hot dense state almost 14 billion years ago, e.g. read and understand What is the evidence for the Big Bang?

 

[Michael]

That is good - like I said the flaw is fairly obvious so that paper is just wrong to people who read it.

It's good for David and bad for you. The "flaw" is your own flaw, not his. He simply pointed it out.

 

[Michael]

That is wrong. Olbers' paradox is a great example of applying simple logic to the universe.

No, it's a great example of the lies that you tell to unsuspecting children/students as the earlier conversation on Olbers's paradox in this very thread clearly demonstrated. Selfsim and sjastro, and you too if you're peddling that same nonsense, came up 200 billion stars and a 100,000 galaxies short of a valid scientific argument. If you believe that nonsense, then explain to us why we don't see every star in our own galaxy in the night sky with our naked eyes RC.

It can be stated simply, e.g. in a "static, infinitely old universe with an infinite number of stars distributed in an infinitely large space" every line of sight will end on a star and so the night sky cannot be dark.

It's not true that our human eyes could ever hope to see every star in our own galaxy, nor every galaxy in our own supercluster. The inverse square law of light and the limitation of human eyesight precludes Olber's lame argument from ever being valid. It's just another great example of pure mainstream nonsense being falsely promoted as "fact".

Or we can do the math and get the same result.

No you can't. Your math came up 268,770 shells short of a valid argument and you forgot to include the inverse square laws in your formulas.

The reasonable conclusion is that one or more of the assumptions is wrong. The universe is finite, not static, not eternal or a combination of them.

No, the "reasonable" assumption is that the limits of human eyesight and the inverse square laws of light apply to human beings on Earth. The 'reasonable" assumption is that Thomas Digges was right all along, and Olber was spewing utter nonsense and you're just spewing the same nonsense.

The "shell of stars" argument is used for the homogeneous universe that we observe at large scales.

You're 268,770 AU shells short of a valid shell argument too! The next four stars are not found inside of a 2AU shell RC.

That allows the argument to be understood by anyone.

Ya, only anyone gullible enough to believe that there are four more stars found 2AU's from Earth!

It's obvious that you never bothered to read this thread or your wouldn't be spewing the same utter nonsense that was obliterated earlier in this same thread. You're 268,770 AU shells, 200 billion stars, and 100,000 galaxies short of a valid scientific argument.

 

[RealityCheck01]

No, it's a ...

A rant about lies which has nothing to do with Oblers' paradox

In astrophysics and physical cosmology, Olbers' paradox, named after the German astronomer Heinrich Wilhelm Olbers (1758–1840), also known as the "dark night sky paradox", is the argument that the darkness of the night sky conflicts with the assumption of an infinite and eternal static universe. The darkness of the night sky is one of the pieces of evidence for a dynamic universe, such as the Big Bang model. In the hypothetical case that the universe is static, homogeneous at a large scale, and populated by an infinite number of stars, then any line of sight from Earth must end at the (very bright) surface of a star and hence the night sky should be completely illuminated and very bright. This contradicts the observed darkness and non-uniformity of the night.[1]

 

[Michael]

A rant about lies which has nothing to do with Oblers' paradox

RC, do yourself a favor and go back and read the whole thread. Your lame Olber's *non* paradox argument came up 268,770 AU shells, 200 billion stars and 100,000 galaxies short of a valid scientific argument. It has nothing to do with a "big bang". Thomas Digges figured out the real reason that the night sky is dark before Olber was even born. The real reason the night sky is dark is because of the inverse square law of light, and the limitations of human eyesight. Period! If Olber was correct we should see the over 200 billion stars in our own galaxy and about 100,000 or so galaxies in our local supercluster. Instead we see less than 10,000 stars, and less than 10 galaxies at night with our naked eyes.

The fact that you still believe that Olber's paradox is a valid argument, or that it supports a big bang theory says volumes about the gullibility of your entire industry and it demonstrates how little you know about light. You simply parrot each other's false and irrational beliefs over and over again without ever thinking for yourselves. That's why you're still stuck in the dark ages of astronomy.

 

[RealityCheck01]

It's good for David and bad for you....

I wrote about the obvious and fatal flaw in David F Crawford's paper that anyone can understand.

26 February 2018: The paper purposely removed the variation in the width of light curves by fitting them to a template light curve with an average width.
My simple example at CosmoQuest

Think about the variation of the height of people with shoe size. Calculate a reference height from a small sample. Measure the shoe sizes and heights of a larger sample. Fit the measured heights to the reference height, i.e. set them to its value. No one would expect that plotting the fitted heights versus shoe size tells us about the variation of the measured heights of people with shoe size.

Crawford's manipulation is more complex - he fits both the width and height of light curves to the template and iterates until he has the best fit.

The CosmoQuest thread suggests it is a lie that any (1 + z) factor is removed in the calibration, a bit because David Crawford could not support his assertion with citations to where it was removed but mostly because time dilation is measured in calibrated type 1a supernova light curves !

 

[RealityCheck01]

RC, do yourself a favor .....

Michael, do yourself a favour and learn what Olbers' paradox is. These is none of this fantasy in it.

 

[Michael]

Michael, do yourself a favour and learn what Olbers' paradox is. These is none of this fantasy in it.

I know exactly what Olber's paradox is, specifically it's a bogus and ridiculously childish argument that has no basis in reality. A dark night sky does *not* support any particular cosmology model. It simply demonstrates the validity of the inverse square laws of light, and the limitation of human eyesight.

If Olber's paradox was a valid argument, we should see about 200 billion more stars in the Milky Way galaxy at night, none of which are expanding away from us. Due to the inverse square laws and the limitation of human eyesight, we see less than 10,000 stars total!

If your "shells" claim was true, then we should have 4 more stars within 2 AU from Earth. Instead, there isn't single extra star found in the next 268,700 "shells".

Olber's paradox is a myth that was "debunked" by Thomas Digges long before Olber was even born.

 

[Michael]

I wrote about the obvious and fatal flaw in David F Crawford's paper that anyone can understand.

It's not a flaw in David's paper, it's a flaw in your SN1A evaluation method!

The CosmoQuest thread suggests it is a lie.....

You have really bad habit of claiming that everything that you personally disagree with is a "lie". It's not.

You simply cheat at debate by asserting dishonest intent where none is warranted, regardless of the topic or the person. Meanwhile you flat out *lied* when you claimed that Scott's solar model predicts "no neutrinos" and excess gamma rays. Your own lies aren't even logically consistent which is why it' *obviously* your problem, not their problem. It's not everyone else that is telling the lies RC, it's the guy looking back at you in the mirror.

 

[RealityCheck01]

Odd that distinguished astronomer and cosmologist, Edward 'Ted' Harrison, devoted an entire chapter of his book 'Cosmology - The Science of the Universe' to the Olber's paradox 'nonsense' (see chapter 24).

"...developments in cosmology have made little difference to the riddle. In a universe of infinite extent, populated everywhere with bright stars, the entire sky should be covered by stars with no separating dark gaps..."
​

Seems like you are the lone voice - the St. John the Baptist, if you will - of cosmology

Nevertheless, I recommend a read of Chapter 24 - it's 22 pages of easy reading - it might have been written for novices and lone voices

Chapter 24 starts

There is a simple and important experiment in cosmology that almost everybody can perform. It consists of gazing at the night sky and noting its state of darkness. When we ask, why is the sky dark at night?

We get the classic "every line of sight must end in a star" and so there cannot be any dark gaps. Thomas Digges wrote the first Copernican analysis but

Digges main contribution was to point out that although the stars are infinitely numerous, yet only a finite number can be seen because ‘‘the greatest part rest by reason of their wonderfull distance invisible unto us.’’ With these words he originated the riddle of the dark night sky. He was the first to realize that dark gaps between the stars call for an explanation. His solution, however, which seemed at the time obvious and acceptable, is incorrect.

(my emphasis added).
On Page 495, is the first mathematical treatment of the paradox where Edmund Halley introduced looking at shells of stars. So in a universe with a uniform distribution of stars, each shall is the same brightness and an infinite number of shells = infinite brightness. The wrong conclusion because the argument does not include that stars obstruct each other. Halley repeated Digges argument and Page 496 states why this is in error

The arguments made by Digges and Halley cannot be true because both neglected the collective light of many stars. In the forest analogy, their solution assumes that distant trees are invisible and we see only foreground trees.

 

[RealityCheck01]

It's not a flaw in David's paper, it's a flaw in your SN1A evaluation method!

Now we have a blatant lie.
5 March 2018: A blatant lie that the obvious fatal flaw in Crawford's paper is a flaw in type 1a supernova light curve calibration.

The blatant part of the lie is that I pointed out the flaw in Crawford's paper.
26 February 2018: The paper by Crawford purposely removed the variation in the width of light curves by fitting them to a template light curve with an average width.

This is not what the calibration does. Sample data is used to create an algorithm that uses the template and several other files to calibrate the light curves.
Welcome to the SALT web page.

SALT (for Spectral Adaptive Lightcurve Template) is a package for Type Ia Supernovae light curve fitting. Its main purpose is to provide distance estimates but it can also be used for photometric redshifts, and spectroscopic + photometric identification (see the References page for links to papers).

SALT2 has 15 template files (averages, variability, error corrections, etc.) of which Crawford only uses the average spectral sequence !

For interested people, The SALT model starts in 2005.
SALT: a spectral adaptive light curve template for type Ia supernovae

We present a new method to parameterize type Ia Supernovae (SN Ia) multi-color light curves. The method was developed in order to analyze the large number of SN Ia multi-color light curves measured in current high-redshift projects. The technique is based on empirically modeling SN Ia luminosity variations as a function of phase, wavelength, a shape parameter, and a color parameter. The model is trained with a sample of well-measured nearby SN Ia and then tested with an independent set of supernovae by building an optimal luminosity distance estimator that combines the supernova rest-frame luminosity, shape parameter, and color reconstructed with the model. The distances we measure using B- and V-band data show a dispersion around the Hubble line comparable or lower than obtained with other methods. With this model, we are able to measure distances using U- and B-band data with a dispersion of 0.16±0.05 around the Hubble line.

As anyone can read the model is trained to calibrate using phase + wavelength + shape + color data.

 

[Michael]

Chapter 24 starts

We get the classic "every line of sight must end in a star" and so there cannot be any dark gaps.

Horse manure. That would only be true if light didn't follow the inverse square law. In reality however the star could be so far away that even though our line of sight ends at that star, that distant star is still far too dim to see with our naked eyes!

How can you ignore the inverse square law and the limitation of human eyesight? We don't even see every star in every distant galaxy, just the ones that are large enough and bright enough for enough of their light to reach Earth!

Thomas Digges wrote the first Copernican analysis but

But nothing. He was right all along, and Olber messed up big time by not considering the inverse square law and the limitation of human eyesight.

(my emphasis added).
On Page 495, is the first mathematical treatment of the paradox where Edmund Halley introduced looking at shells of stars. So in a universe with a uniform distribution of stars,

Bzzt! There is no uniform distribution of stars from the perspective of Earth. The first star is 1 AU from our planet, whereas the next one is over 268,770 shells away! There aren't four more stars at 2AU from Earth so that claim is absurdly false.

The issue has nothing to do with stars blocking anything, it has to do with distance, the inverse square law, an the limitation of human eyesight.

Man, what a lame argument. I see now why we're stuck in the dark ages of astronomy. You guys just repeat the same falsified dogma over and over again and you never bother to think for yourselves. You're also clueless about the inverse square laws of light.

In the forest analogy, their solution assumes that distant trees are invisible and we see only foreground trees.

Their solution is entirely accurate due to the inverse square law of light, which is why we see less than 10,000 stars out of the 200+ *billion* stars in our own galaxy! Gah!

 

[Michael]

Now we have a blatant lie.

No, we have another example of a *disagreement*, and another example of you blatantly cheating at debate by implying dishonest intent where none is necessarily warranted or required. Tonight alone you accused me, David and Justa of all "lying" simply because you disagreed with them. That's just childish egotistical nonsense. You aren't the sole decider of 'truth'. In fact we know for a fact from your misrepresentations of Scott's solar model that it is *you* that has a serious problem with not telling the truth.

From the OP at Cosmoquest:

I have just had a paper published in Open Astronomy (2017) 19,111 (arXiv:1711.11237) with the title "A problem with the analysis of type Ia supernovae"
There are three major independent results in the paper:

A: the standard analysis of the light curves from type Ia supernovae (SALT2) has a major flaw. Although it correctly corrects the light curve for intrinsic variations as a function of wavelength it unfortunately removes all the cosmological information from the light curve. In particular the calibrated widths well not show time dilation as a function of redshift. Note that this width data provides the dominant observational evidence for dark energy which means that the existence of dark energy is suspect.

B: An analysis of the original raw light curve data shows no.time dilation and is fully compatible with a static universe.

C: The peak magnitudes are consistent with the distance modulus of an existing cosmological model for a static universe.

This is not what the calibration does. Sample data is used to create an algorithm that uses the template and several other files to calibrate the light curves.
Welcome to the SALT web page.

You're essentially "whining" because Crawford specifically and intentionally chose to *not* analyze the data in the same exact way! That was never his intent.

SALT2 has 15 template files (averages, variability, error corrections, etc.) of which Crawford only uses the average spectral sequence !

For interested people, The SALT model starts in 2005.
SALT: a spectral adaptive light curve template for type Ia supernovae

As anyone can read the model is trained to calibrate using phase + wavelength + shape + color data.

So what? Crawford *intentionally* did *not* process the data the same exact way as SALT 2 and he stated that fact in his paper. The intent of not using the SALT 2 templates and methods was to show that the SALT 2 methods were *flawed*.

The raw SNe observations were analyzed without using the SALT2 method. For each SNe the fitted parameters were the peak flux density, the epoch of the peak flux density and the relative width of the light curve. This fitting was done using the reference light curve provided by Goldhaber et al. [10]. Then the two scale parameters, the peak flux density and the width, (and the epoch of peak flux density) were determined for each filter. Since these scale parameters are orthogonal they can be determined by separate least squares analysis. For the peak flux density this done by minimizing the average square of the difference between the observed flux density at each epoch and the height of the reference curve at that epoch multiplied by the peak flux density and divided by the uncertainty in the flux density. In other words a minimum χ2 analysis.

Emphasis mine.

 

[RealityCheck01]

That would only be true if light didn't follow the inverse square law...

9 March 2018: Continued ignorant denial of Oblers' paradox.
The mathematical derivation of Oblers' paradox includes the inverse square law of light !
It is that inverse square law + the square law of the surface area of a shell that gives a constant brightness for each shell.
It is the finite size of light sources that blocks light and stops an infinite number of shells from making the night sky infinitely bright.

 

[RealityCheck01]

No, we have another example of a *disagreement*, ....

A lie about what I wrote is not a disagreement.
5 March 2018: A blatant lie that the obvious fatal flaw in Crawford's paper is a flaw in type 1a supernova light curve calibration.

The blatant part of the lie remains that I pointed out the flaw in Crawford's paper.
26 February 2018: The paper by Crawford purposely removed the variation in the width of light curves by fitting them to a template light curve with an average width.

Welcome to the SALT web page. and its 15 SALT2 template files that show that the "Average spectral sequence" file is not the only file used to calibrate light curves. That makes any analysis of one of the files another fatal error because he does not know what the other files do. Look at the next file:
salt2_template_1.dat Spectral sequence of 1st variability component

Ignorance about Crawford's paper - the "debunking" of SALT2 is not his flawed analysis of light curves. The "debunking" is in the previous section where there is an error about what the salt2_template_0.dat (Average spectral sequence) contains. This file contains an average spectral sequence over many type 1a supernova observed by multiple telescopes. Crawford has an analysis that starts with one telescope observing one type 1a supernova.

 

[Michael]

A lie about what I wrote is not a disagreement.

Oy Vey. Your liar, liar pants on fire routine is *childish* RC. You've never told the truth about *any* model that you disagree with or any paper you disagree with. You misrepresent them, just like you're misrepresenting David's paper. Your lie claims are pure projection.

5 March 2018: A blatant lie that the obvious fatal flaw in Crawford's paper is a flaw in type 1a supernova light curve calibration.

The blatant part of the lie remains that I pointed out the flaw in Crawford's paper.

What you pointed out isn't a flaw in his paper, it's a flaw in the SN1A analysis techniques which David pointed out!

26 February 2018: The paper by Crawford purposely removed the variation in the width of light curves by fitting them to a template light curve with an average width.

He didn't "remove" anything. You're the one removing data, which is the whole point of his paper.

The rest of your response is just a blatant rehash of the same issue. You're falsely accusing him of doing what SALT2 is doing with their own data by fitting it to a series of template files which strip out important information.

The worst part of all this is the fact that more recent and comprehensive SN1A studies do not support your claim of 'discovery" of dark energy in the first place!

Forbes Welcome

Not only are your SN1A analysis methods questionable to start with, the results are not conclusive either!