Recreating a slice of the universe

Aug 15, 2012
This still frame is taken from the Arepo-generated animation shown above. It demonstrates Arepo's key ability to produce realistic spiral galaxies. Previous simulations tended to yield blobby galaxies lacking distinct spiral structure. Credit: CfA/UCSD/HITS/M. Vogelsberger (CfA) & V. Springel (HITS)

(Phys.org) -- Scientists at the Harvard-Smithsonian Center for Astrophysics (CfA) and their colleagues at the Heidelberg Institute for Theoretical Studies (HITS) have invented a new computational approach that can accurately follow the birth and evolution of thousands of galaxies over billions of years. For the first time it is now possible to build a universe from scratch that brims with galaxies like we observe around us.

"We've created the full variety of galaxies we see in the local universe," said Mark Vogelsberger (CfA).

Our cosmic neighborhood is littered with majestic spiral galaxies like Andromeda, the Pinwheel, and the Whirlpool. are common, but previous simulations had trouble creating them. Instead, they produced lots of blobby galaxies clumped into balls, without the broad disks and outstretched arms of a typical spiral.

The new software, called Arepo, solves this problem. Created by Volker Springel (HITS), Arepo generates a full-fledged simulation of the universe, taking as input only the observed of the and evolving forward in time for 14 billion years.

This video is not supported by your browser at this time.
This computer animation, created using new software called Arepo, simulates 9 billion years of cosmic history. Arepo can accurately follow the birth and evolution of thousands of galaxies over billions of years. Arepo generates the full variety of galaxies seen locally, including majestic spirals like the Milky Way and Andromeda. Credit: CfA/UCSD/HITS/M. Vogelsberger (CfA) & V. Springel (HITS)

"We took all the advantages of previous codes and removed the disadvantages," explained Springel.

"Our simulations improve over previous ones as much as the Giant Magellan Telescope will improve upon any telescope that exists now," said Debora Sijacki (CfA).

(When completed later this decade, the Giant Magellan Telescope's 24.5-meter aperture will make it the largest telescope in the world.)

One of Arepo's key advantages is the geometry it uses. Previous simulations divided space into a bunch of cubes of fixed size and shape. Arepo uses a grid that flexes and moves in space to match the motions of the underlying gas, stars, dark matter, and dark energy.

The simulations ran on Harvard's Odyssey high-performance supercomputer, using in total 1024 . This fast machine allowed the scientists to compress 14 billion years into only a few months - an endeavor that would have kept a desktop computer busy for hundreds of years!

The team's future goals include simulating much larger volumes of the universe at unprecedented resolution, thus creating the largest and most realistic model of the universe ever made.

Explore further: Staying warm: The hot gas in clusters of galaxies

More information: The team consists of Mark Vogelsberger (CfA), Debora Sijacki (CfA), Dusan Keres (CfA/UCSD), Paul Torrey (CfA), Volker Springel (HITS), and Lars Hernquist (CfA). Their work is described in three papers accepted for publication in the Monthly Notices of the Royal Astronomical Society. Those papers can be found online at arxiv.org/abs/1109.1281 , arxiv.org/abs/1109.3468 , and arxiv.org/abs/1109.4638 .

Related Stories

Strange new 'species' of ultra-red galaxy discovered

Dec 01, 2011

In the distant reaches of the universe, almost 13 billion light-years from Earth, a strange species of galaxy lay hidden. Cloaked in dust and dimmed by the intervening distance, even the Hubble Space Telescope ...

Cosmic Dance Helps Galaxies Lose Weight

Jul 29, 2009

(PhysOrg.com) -- A study published this week in the journal Nature offers an explanation for the origin of dwarf spheroidal galaxies. The research may settle an outstanding puzzle in unders ...

Astronomers release the largest ever 3-D map of the sky

Aug 08, 2012

The Sloan Digital Sky Survey III (SDSS-III) has released the largest three-dimensional map of massive galaxies and distant black holes ever created. The new map pinpoints the locations and distances of over ...

Mysterious red galaxies

Dec 12, 2011

(PhysOrg.com) -- Perhaps the most astonishing and revolutionary discovery in cosmology was Edwin Hubble's observation that galaxies are moving away from us with velocities that are proportional to their distances. ...

Recommended for you

Staying warm: The hot gas in clusters of galaxies

21 hours ago

Most galaxies lie in clusters, groupings of a few to many thousands of galaxies. Our Milky Way galaxy itself is a member of the "Local Group," a band of about fifty galaxies whose other large member is the ...

A colorful gathering of middle-aged stars

Nov 26, 2014

NGC 3532 is a bright open cluster located some 1300 light-years away in the constellation of Carina(The Keel of the ship Argo). It is informally known as the Wishing Well Cluster, as it resembles scattered ...

User comments : 8

Adjust slider to filter visible comments by rank

Display comments: newest first

Kedas
not rated yet Aug 15, 2012
Only a bit more detail and they will be able to see the lottery numbers of next week ;)
hemitite
not rated yet Aug 15, 2012
A billion here, a billion there - pretty soon you've got some real time on your hands!
GSwift7
3 / 5 (2) Aug 15, 2012
Only a bit more detail and they will be able to see the lottery numbers of next week ;)


Or Einstein may have been wrong, and God does play dice.

taking as input only the observed afterglow of the Big Bang and evolving forward


I wonder how true that really is. Since it's not possible to actually simulate everything, there must be parameterizations for things like turbulence in weather models. So, technically, since such parameterizations would be based on observations, they 'could' be counted as input too.
nevermark
3.7 / 5 (3) Aug 15, 2012
So, technically, since such parameterizations would be based on observations, they 'could' be counted as input too.


Well in that kind of thinking everything is always input. Programs are inputs to computers.

But the distinction usually taken is that the model is whatever stays the same throughout a simulation (i.e. constant parameters, the repeatedly applied mathematical relationships from each instant to the next, etc.) and the starting state and ending states are considered the input and output.

In a case like this, the input data is many orders more information than the model, and the output even larger, which makes the success of the model significant.
Torbjorn_Larsson_OM
5 / 5 (4) Aug 15, 2012
@ GSwift7:

I think they refer to this, from the 1st paper:

"An important reason for todays reliability of CDM predictions is that the initial conditions are unambiguously specified in the CDM model, with parameters that are tightly constrained by CMB observations (Komatsu et al. 2011). Moreover, the computational problem is well-posed and comparatively simple, with equations of motion that for DM involve only gravity. Efficient new algorithms and the rapid growth of computing power over the few last decades have allowed ever more detailed theoretical predictions for the DM distribution."

So it is a very well constrained problem with well defined physics and initial conditions. The problem lies in the simulations. Parameterizations should be considered as additional constraints (or physics), the trick is to know them and suppress their effects.
TimESimmons
1 / 5 (5) Aug 16, 2012
Once again they create spirals because the simulation uses gravitational softening. If they run it again without (or with much reduced) gravitational softening it will not produce spirals. I'll email them that challenge.

In the real universe spirals are produced by AGM:-
http://www.presto...ndex.htm
roboferret
not rated yet Aug 16, 2012
Once again they create spirals because the simulation uses gravitational softening. If they run it again without (or with much reduced) gravitational softening it will not produce spirals. I'll email them that challenge.

In the real universe spirals are produced by AGM:-
http://www.presto...ndex.htm


Agreed, almost every mystery in science can be accounted for by AGM (Alien Ghost Monsters).
GSwift7
5 / 5 (1) Aug 16, 2012
Well in that kind of thinking everything is always input. Programs are inputs to computers.

But the distinction usually taken is that the model is whatever stays the same throughout a simulation


Well, since we don't have intimate knowledge of methodology, my question still stands. You are making an assumption, unless you know for certain that there aren't other inputs which strongly affect the results. Posibilities that come to mind as additional inputs include proportions of anti/matter, cosmic abundance of elements, ratio of 'normal matter' to dark matter, chandrasekhar's limit, etc. You can call that stuff "constants" under some circumstances, but depending on how the program uses them, they could also be called input variables.

I know it's hard to imagine a reporter trying to sensationalize a story or exagerate something, but it IS possible that it happens once in a while. ;P

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.