# Number of Voids in the (observable) Universe

## Main Question or Discussion Point

Hi,

Does maybe someone know if there was ever the attempt to count the number
of voids in the (observable) universe?

I assume that is not the case, but would be very interested in any work that
gives some estimation.

Many thanks,
Martin

## Answers and Replies

phinds
Gold Member
2019 Award
Hi,

Does maybe someone know if there was ever the attempt to count the number
of voids
in the (observable) universe?

I assume that is not the case, but would be very interested in any work that
gives some estimation.

Many thanks,
Martin
What does that even mean? Why would there not just be one void, interspersed here and there (at VERY long distances) with galaxies and galactic clusters, and even strings of clusters?

Hi Phinds,

thanks for your reply and sorry for not expressing my question correctly enough!

If you consider a voids of specific size, like here

http://en.wikipedia.org/wiki/List_of_voids

Is there any study that says:

There are roughly about N voids with diameter M in the observable universe.

I searched the net and found... nothing. Also I assume noone, except me is interested in that question ;-)

Thanks,
Martin

jimgraber
Gold Member
Answer: Roughly a million voids
The power spectrum of the cosmic microwave background shows a big bump at around half a degree or slightlly larger. According to the (broadly accepted) acoustic theory, clumps and voids should occur at roughly this scale, on average. This scale is approximately 100 megaparsecs today. You wil notice that the voids in your wikipedia list are mostly within a factor of a few bigger or smaller than one hundred parsecs. One parsec is about three and a quarter light years, so a typical void is about 325 million light years. the observable universe has a radius of about 13.7 billion light years, so, the typical void is 1/40th or 1/50th of the radius or around 1/100th of the diameter of the observable universe. So for a cubical universe, there should be approximately 100^3 or about a million voids. For a spherical universe (pi r^3 and all that), you get a slightly smaller result, but we are nowhere near that accuracy. Also, a more correct calculation would do a more accurate treatment of the cosmic expansion, etc.

But roughly a million expected voids is correct to an order of magnitude. We've known this since WMAP, which lead to Nobel prizes for this discovery.

jimgraber
Gold Member

Take a look at the WMAP picture here in Wikipedia. The smallest blue dots are typical voids at the edge of the observable universe. They should be about 1/360 th of the height or 1/720 th of the width.
That one big blue splotch near the middle is so big people worry about it.

jimgraber
Gold Member
phinds is right that the voids do run together and are not well separated. Roughly, the blobs (superclusters) are stronger that the filaments, which are stronger than the walls. Most of the walls are "very leaky" or not very well defined, with some famous exceptions. Some of the nearby voids are amazingly empty.

Hi jimgraber,

thank you very much for your precise answer!

That helped me a lot. I was playing with the *idea* that it could
be around roughly a million, but I could not find any evidence.

Many thanks again,
Martin

Chronos
Gold Member
The void and cluster issue is a matter of definition - what constitutes a void, or cluster? Is a hundred enough, or too few. This is a perception issue. We can characterize the universe at large, but, not by voids and clusters. That is irrelevant, IMO.

jimgraber
Gold Member
I just thought of another important point. The far away voids aren't very empty. Just slight underdensities, surrounded by slight overdensities. As time goes on, the voids empty out, and the clusters contract. So the closer voids are much emptier than the far away voids.

phinds
Gold Member
2019 Award
I just thought of another important point. The far away voids aren't very empty. Just slight underdensities, surrounded by slight overdensities. As time goes on, the voids empty out, and the clusters contract. So the closer voids are much emptier than the far away voids.
If I understand what you are saying, it is nonsensical. EVERY void is very far away from some frame of reference. Are you trying to make us the center of the universe?

BillSaltLake
Gold Member
What might be informative here is a plot of observations (or a plot derived from the millennium simulation) which shows the distribution of energy density. The x axis would be space volume, with the least dense on the left ("x"=0) and the most dense on the right ("x"=1 or 100%-- this is a histogram format). The y axis is energy density, although it would not include "dark energy". In early times, the plot would just be a horizontal line (= uniform density). Then as clumping occurs, the left side drops while the right side of the curve rises.
I've been looking for a set of curves (for several values of z) like this from the millennium simulation, but haven't found it yet. With this set of plots, you could get a feel of at least what fraction of space is void. In other words, what fraction of space contains < 1% (for example) of the average density, at (for example) z=3.

Last edited:
jimgraber
Gold Member
"If I understand what you are saying, it is nonsensical. EVERY void is very far away from some frame of reference. Are you trying to make us the center of the universe?"

Well, yes. We are the center of our *observable* universe. Far away is also backwards in time, and the voids we see at the edge of the (our) observable universe, ie in the CMB underdensities, are very "young". When they are as old as the nearby voids, they will look similar to the nearby voids, but we won't see that for another thirteen billion years or so.
Best, Jim Graber

phinds
Gold Member
2019 Award
"If I understand what you are saying, it is nonsensical. EVERY void is very far away from some frame of reference. Are you trying to make us the center of the universe?"

Well, yes. We are the center of our *observable* universe. Far away is also backwards in time, and the voids we see at the edge of the (our) observable universe, ie in the CMB underdensities, are very "young". When they are as old as the nearby voids, they will look similar to the nearby voids, but we won't see that for another thirteen billion years or so.
Best, Jim Graber
Ah ... now I get what you mean.

By the way, the quote button is very easy to use.

I just thought of another important point. The far away voids aren't very empty. Just slight underdensities, surrounded by slight overdensities. As time goes on, the voids empty out, and the clusters contract. So the closer voids are much emptier than the far away voids.

Have you considered the Eridanus Supervoid or Great Void?

I searched the net and found... nothing. Also I assume noone, except me is interested in that question ;-)
It's actually out there just in astrophysics language.

The term that astrophysicists use to ask that question is "power spectrum".

See figure 6

http://universe-review.ca/R05-04-powerspectrum.htm

What that diagram says is that the size of a void is about 1 degree, which gives you 1/60th of the radius of the universe which gives you the numbers that jimgraber gives you.

Also these numbers don't tell you how "deep" the voids are.

Last edited:
I've been looking for a set of curves (for several values of z) like this from the millennium simulation, but haven't found it yet. With this set of plots, you could get a feel of at least what fraction of space is void. In other words, what fraction of space contains < 1% (for example) of the average density, at (for example) z=3.
I think this might be what you are looking for.

http://www.lanl.gov/projects/cosmology/sf09/carlson_sf09_pt.pdf

Also structure formation near the big bang is something that seems to be simple. Structure formation now is something that we don't totally quite understand.

One non-trivial part of cosmology is how do describe the distribution of the universe. The language that people end up using is identical to the language people use to describe sound.