New model predicts the size of the cosmos

[crossfire]

Speaking of assumptions, here's a new paper that blows the standard assumptions about the curvature of the universe out of the water.

Almost all work on predicting the actual physical size of the cosmos is based on an assumption about the curvature of the universe--whether it is closed (curvature < 0), flat (curvature = 0), or open (curvature > 0

Comments

[bord_du_rasoir]

If the visible universe is 90 billion light years in diameter. Then, this estimate places the universe at 250 x 90 billion or 22.5 trillion light years in diameter

Local spatial geometry and curvature

[crossfire]

Curvature is basically is a way of saying whether or not the Pythagorean Theorem applies to spatial coordinates. A curvature of 1 means the theorem applies, while positive or negative curvatures indicate the theorem does not apply (though the deviations are only detectable with triangles of cosmic sizes).

The typical illustrations you'll see for curvature all involve the projection of three-dimensional space onto two dimensional surfaces. That's convenient because everyone can easily imagine it, but it breaks down when you start getting into the details.

Um. Here, believe it or not Wikipedia has a pretty good entry on this:
http://en.wikipedia.org/wiki/Shape_of_the_Universe

[goseaward]

This compares to the maximum diameter the universe would be if the universe were not expanding: 28 billion light years in diameter...That's the maximum visible diameter of the universe if the universe isn't expanding, not the maximum physical diameter. In the case where the universe isn't expanding, it can be infinite if you want: that's just how big it is, set by some initial conditions we don't understand. So the comparison to make is the 28 billion light years to the 90 billion light years

[goseaward]

HOLY TEAL DEER, BATMAN. o.O

[goseaward]

I dunno, I'm a little skeptical. "We ruled out part of our parameter space, so we got better constraints" is a philosophical argument, not a scientific one, imho.

I shouldn't say that--the paper is way more interesting to me than the abstract; they're basically comparing the results you get by assuming several loose things about the universe to the results you get by assuming stricter things about curvature and nothing about anything else (apart from dark energy), which is a fairly interesting exercise, and then they get numbers. And, yeah, I don't really trust numbers in cosmology that aren't densities or temperatures (and I'm not sure about the densities), but it's an interesting illustration of concept.

Actually the other really interesting thing to me is that the ~250 times the size of the Universe is a lower bound! I haven't done much with sizes of the universe (my professors are much more fond of the ages of the universe) and now I want to go poke at the equations and see why that is.

[crossfire]

I wonder if there may not be other applications of this particular methodology. I'd never thought about using Bayesian analysis in this context, I think it's a really clever idea.

As for the philosophical aspect, have you read Vardanyan's (et al) other paper, A model comparison perspective on the curvature of the Universe?

[layweed]

Between this and Kepler finding an Earth-sized planet, tomorrow's Astrobio class is going to be all sorts of awesome!