Why didnt hubble discover that universe is accelerating?

[humanist rho]

In other words why high redshift data is necessary for estimating the acceleration of universe?

 

[Drakkith]

Edwin Hubble did discover this.

 

[Andrew Mason]

Edwin Hubble did discover this.

You may be confusing Hubble's discovery that the universe was expanding with the much more recent discovery that the rate of expansion is increasing.

AM

 

[FtlIsAwesome]

In other words why high redshift data is necessary for estimating the acceleration of universe?

As the speed of an object away from us increases, relativistic effects cause its light to redshift. From the amount of redshift, we can calculate the object's speed.

 

[Drakkith]

You may be confusing Hubble's discovery that the universe was expanding with the much more recent discovery that the rate of expansion is increasing.

AM

Oh, my mistake. I understand the question now. I'm assuming that it took years of accumulated data to show that the acceleration is increasing?

 

[marcus]

Oh, my mistake. I understand the question now. I'm assuming that it took years of accumulated data to show that the acceleration is increasing?

It's really interesting, how they figured out there was acceleration. It's a good question to ask.
There's probably a good explanation on the web, of how it was done, maybe even at Wikipedia (not always the best but sometimes quite good.)

The technical explantation (very briefly) is curve-fitting. There is an equation called after Alex Friedmann that he derived from Einstein's main GR equation. This Friedmann equation has a few numbers (parameters) you plug in and you get a curve that describes the size a(t) of a generic piece of the U evolving over time while obeying the GR law of gravity.

So cosmologists (since they tend to trust Einstein GR and therefore provisionally rely on Friedmann equation) have adjusted the parameters to get the best fit to the masses observational data. Like galaxy counts at different redshifts. That best-fit model with the best-fit parameters gives you a calculator that can convert observed redshift to numbers you want to know like distance then (when the light was emitted) and distance now (when it is received) and light travel time and estimated age of U, and so on.

One of the original parameters that Einstein put in was one called Lambda that causes a(t) growth to speed up and fights the tendency of other gravity to slow it down. That parameter Lambda carries over to the simple Friedmann equation that cosmologists use.
For many years no effect of Lambda was seen, everything fitted the data OK with Lambda set equal to zero.

The kicker is in the calculator that converts redshifts to distances. In 1998 two teams reported that after some years of studying a type of supernova there was a discrepancy. The explosions were actually slightly farther away than you would calculate, from the redshift, as long as you kept the parameter Lambda = 0. To get the right answers you had to change Lambda to make it very slightly positive. Leading to a very gradual acceleration.

We always knew (for years) that a(t) was increasing. The scalefactor curve sloped up. The time-rate-of-change a'(t) was positive. But we thought the growth of a(t) was slowing down, which is what happens if Lambda is zero. So we thought the slope got less steep with time.
But the supernova studies report forced people to conclude that the slope of the a(t) curve was getting steeper with time.

All this came about because of an observed technical discrepancy in the relation of redshift z to distance.

You might want to actually check out a cosmologists calculator that converts z to distance:
Google "wright calculator" or "cosmo calculator". Wright is a UCLA cosmologist who has one on-line that you can use.
=======================

There is a professor named Siobhan Morgan who has an alternative calculator which I like very much partly because when you first get there it is set to pre-1998 parameter values! And you have to manually adjust the parameters. It is easier than it sounds. Where it says 1, you type in .27 and where it says 0 (for Lambda) you type in .73. Prof. Morgan forces you to make the adjustments that astronomers had to make in 1998. It is an educational trick of hers. Cute, but effective.
If you want to try that one, google "cosmos calculator". I also have the link in my signature at the end of the post.

Ned Wright doesn't make you do that. He already has the numbers .27 and .73 in the proper boxes. You can change them if you want.

So you can pretend you are observing a Type Ia supernova and the redshift of the light is, say 0.2 (wavelengths 20% longer) so z = 0.2
And you plug that into Morgan's, with the wrong parameters, and you get a distance which is just a bit short.
Then you plug z=0.2 into Wright's, or you correct the parameters, and you get the real distance that was observed.
It is just a little farther than what was predicted with the older pre-1998 model.

There is more to story, like how do you tell the distance to a Type Ia supernova, and maybe other people will reply to your post.

 

[Andrew Mason]

...I'm assuming that it took years of accumulated data to show that the acceleration is increasing?

I think the main reason was improvement in technology. http://www.pnas.org/content/96/8/4224.full" is particularly good in explaining how it was discovered that the expansion rate was increasing (at least that is how the data is being interpreted).

 

[twofish-quant]

Oh, my mistake. I understand the question now. I'm assuming that it took years of accumulated data to show that the acceleration is increasing?

No. The problem is that you need two things

a) an object that you know has a standard "real" brightness
b) enough of those objects at long distances so that you can measure the distance

It took a while to figure out that supernova Ia are all the same "real" brightness, and then you needed advances in telescope and CCD technology to see really dim supernova that are at far distances. CCD is a fancy way of saying "digital camera."

One thing that happened in the 1990's was that advances in technology made it easier to create large sensitive telescope. A lot of the technology that is used in telescope detectors is basically the same as the things you use in your digital camera. The only real difference is that astronomical CCD's are cooled by liquid nitrogen to remove "heat noise."

So what you could do starting around 1995, is to take lots of digital pictures of the sky, download them to your computer, and then run programs to automatically look for dim supernova. You could also use cheap microcomputers to more quickly design and maintain big telescopes.

 

[humanist rho]

thank u all.
but I've some more questions.
will analysing hubbles data with recent technology give an accelerating universe?
is it possible to have a universe which is accelerating in large redshifts and decelerating at small redshifts?

 

[killinchy]

This is my first post, so please be gentle. I am also a Chemist, so please be doubly gentle.

My question is about a plot found in the paper referred to earlier by Andrew Mason

http://www.pnas.org/content/96/8/4224.full"

The y-axis of the lower panel is delta(m-M)

The paper says, "The lower panel, with the slope caused by the inverse square law taken out, shows the difference between the predictions more clearly and shows why a model with ΩΛ > 0 is favored. with the slope caused by the inverse square law taken out, shows the difference between the predictions more clearly and shows why a model with ΩΛ > 0 is favored."

so far, so good

I know what m is the apparent magnitude and M is the absolute magnitude

delta(m-M) = (m₁-M₁)-(m₂-M₂)

"1" refers to the magnitudes as measured, but what does "2" refer to ?

thanks

John

 

[Drakkith]

I see now. Thanks for the replies all.

 

[bapowell]

thank u all.
but I've some more questions.
will analysing hubbles data with recent technology give an accelerating universe?
is it possible to have a universe which is accelerating in large redshifts and decelerating at small redshifts?

It's not a matter of analyzing data with better technology. The statistical techniques that one would use to analyze Hubble's data are essentially the same as those that one would use to analyze SN data. The key was pointed out by 2-fish: you need a larger "lever-arm" to get sufficient data and statistics to ascertain that the universe is accelerating. This means going out to much greater distances than Hubble was capable of. Hubble used Cepheid variable stars as standard candles, but these become less accurate indicators of distance the farther out they are. Hubble's data was also adversely affected by the fact that the motion of nearby galaxies are more strongly affected by their peculiar velocities -- local motion that deviates from the cosmological expansion -- than more distant galaxies. No modern-day analysis technique would be able to mollify these sources of error. In fact, while more modern instrumentation might record even more accurate luminosity distances and redshifts than Hubble was capable of, the most important factor is still the distance scale. More recent investigations have considered type Ia supernovae because they are bright and enable an accurate determination to greater distance (the 1998 data of Riess et al spans a distance about 700 times greater than Hubble's data!)

If the universe were decelerating locally and accelerating more distantly, than this would indicate that we, being at the center of this rather interesting phenomenon, would be something special. The cosmological principle states that there are no such privileged observers -- the universe expands isotropically and homogeneously for *all* observers.

 

[humanist rho]

If the universe were decelerating locally and accelerating more distantly, than this would indicate that we, being at the center of this rather interesting phenomenon, would be something special. The cosmological principle states that there are no such privileged observers -- the universe expands isotropically and homogeneously for *all* observers.

didn't the large redshift and small redshift imply near ad distant times?
then what is the difficulty in assuming a universe which is accelerating at earlier times and decelerating in latter times??

 

[phinds]

This is my first post, so please be gentle. I am also a Chemist, so please be doubly gentle.

I once heard it said that the only use physicists have for chemists is to grab them by the ankles and use them to bash the crap out of an engineer.

I'm an engineer, so I think it's OK for me to pass the story on without meaning any insult.

 

[Drakkith]

I once heard it said that the only use physicists have for chemists is to grab them by the ankles and use them to bash the crap out of an engineer.

I'm an engineer, so I think it's OK for me to pass the story on without meaning any insult.

Lol.

 

[Redbelly98]

Didn't the acceleration of the universe's expansion come from CMB data taken by the COBE satellite in the 1990's? Hubble died in 1953, before the CMB was even discovered.

 

[bapowell]

Didn't the acceleration of the universe's expansion come from CMB data taken by the COBE satellite in the 1990's? Hubble died in 1953, before the CMB was even discovered.

No. The first indication that the present-day universe was undergoing accelerated expansion came from observations of supernovae redshifts in 1998 by Riess et al:

http://arxiv.org/abs/astro-ph/9805201

and confirmed in 1999 by Perlmutter et al:

http://arxiv.org/abs/astro-ph/9812133

 

[Redbelly98]

Ah, okay. Thanks for clearing that up!

 

[Chalnoth]

No. The first indication that the present-day universe was undergoing accelerated expansion came from observations of supernovae redshifts in 1998 by Riess et al:

http://arxiv.org/abs/astro-ph/9805201

and confirmed in 1999 by Perlmutter et al:

http://arxiv.org/abs/astro-ph/9812133

Well, technically Redbelly98 is mostly right here, though we actually had to combine the COBE data with other experiments to get a handle on the curvature of the universe from the CMB. Still, at the time, the data was exceptionally noisy, so a lot of people really weren't very confident about it.

See here, for example, published in 1993: http://adsabs.harvard.edu/abs/1993ApJ...413...1K

The breakthrough of the supernova data wasn't so much in terms of discovery, but in terms of providing a result stark enough that few could actually doubt that our universe was accelerating. The data had been pointing in that direction for a while, however.