Universe evolution recreated in lab

6

An international team of researchers has created the most complete visual simulation of how the Universe evolved.

The computer model shows how the first galaxies formed around clumps of a mysterious, invisible substance called dark matter.

It is the first time that the Universe has been modelled so extensively and to such great resolution.

The research has been published in the journal Nature.

The simulation will provide a test bed for emerging theories of what the Universe is made of and what makes it tick.

One of the world's leading authorities on galaxy formation, Professor Richard Ellis of the California Institute of Technology (Caltech) in Pasadena, described the simulation as "fabulous".

"Now we can get to grips with how stars and galaxies form and relate it to dark matter," he told BBC News.

The computer model draws on the theories of Professor Carlos Frenk of Durham University, UK, who said he was "pleased" that a computer model should come up with such a good result assuming that it began with dark matter.

"You can make stars and galaxies that look like the real thing. But it is the dark matter that is calling the shots".

Cosmologists have been creating computer models of how the Universe evolved for more than 20 years. It involves entering details of what the Universe was like shortly after the Big Bang, developing a computer program which encapsulates the main theories of cosmology and then letting the programme run.

The simulated Universe that comes out at the other end is usually a very rough approximation of what astronomers really see.The latest simulation, however, comes up with the Universe that is strikingly like the real one.

Immense computing power has been used to recreate this virtual Universe. It would take a normal laptop nearly 2,000 years to run the simulation. However, using state-of-the-art supercomputers and clever software called Arepo, researchers were able to crunch the numbers in three months.

Written By: Pallab Ghosh
continue to source article at bbc.com

6 COMMENTS

  1. An international team of researchers has created the most complete visual simulation of how the Universe evolved.

    “A picture is worth a thousand words” as the old saying goes, and it is certainly so, where computer animations are used to explain complex issues which are difficult to understand.

    We need to be careful that the animation does actually match reality as far as possible, and that the limitations of this method of presentation is also made clear, but it provides a wonderful tool for looking at the complexities of nature.

  2. “You can make stars and galaxies that look like the real thing. But it is the dark matter that is calling the shots”.

    Ah! A new theology is formed to go with the new simulation! Obviously the Dark Matter must be God!

  3. Dare I say it . . . . ?

    I am not setting myself up as being more educated or intelligent than any of the people involved in this computer simulation, so please do not misunderstand my post. However, the article raises several questions.

    It shows that certain assumptions were fed into the computer software in order to produce the simulations that resulted. (For example, it states that it is assumed that the universe began with dark matter.) The article then says that the result is remarkably similar to what we actually see.

    In addition to the question of whether the assumptions are correct, surely if one feeds in current assumptions about how universe started and how the present state of the universe got here, and then models the that development, isn’t it obvious that the resulting simulation will be similar to what we see because we have assumed both the starting point and the development model? In other words: all we’ve got is an animated version of what some people think. How do we distinguish between what we can actually know and what is merely assumed (albeit an educated assumption) and does the article make it clear that is a distinction between the two? Also, how do you prove that your assumptions are actually what actually happened at the actual beginning?

    I therefore agree with Alan4discussion at #1 that “We need to be careful that the animation does actually match reality as far as possible, and that the limitations of this method of presentation is also made clear…”

    What beats me is how anyone will ever know if any simulation about the distant past is an accurate reflection of what actually happened when no-one was there to confirm the proposed theory or say otherwise. This or any other simulation could be wildly off the mark but we would never know for sure – actually. So, does this simulation really show us anything?

    Still, it must have been a fun piece of software to play with.

    On a different tack: I’ve heard Pallab Ghosh many times on BBC Radio and I’ve never felt that he comes over in an authoritative way in his science reports. How do people how know of him rate him as a suitably qualified person for the BBC to appoint as a science correspondent? (This is nothing to do with the content of the article or the research discussed therein. It is purely an opinion about the reporter.)

    • In reply to #3 by Lonevoice:

      Dare I say it . . . . ?

      I am not setting myself up as being more educated or intelligent than any of the people involved in this computer simulation, so please do not misunderstand my post. However, the article raises several questions.

      It shows that certain assumptions were fed into the computer…

      It’s a problem you have with any computer simulation. If your model is crap so is your simulation.

      But that doesn’t mean that simulation is useless or that you can’t learn things from it. It can happen that you plug in the model and some data and find that what you expected to happen really doesn’t. That happens a lot with simulations, it’s why they are useful. Or you can play around with random variables and run many simulations. And that way you can find that the simulated behavior may have a lot of variability. E.g. you run the exact same model two times and you don’t change anything in the actual model but part of the model includes variables that are determined by random number generators and you may find that the results are wildly different because of interactions you hadn’t thought about among the various random variables. That’s essentially the definition of a chaotic system, one where a minor tweak in what seem like insignificant random variables can cause major differences in the results.

      Getting back to your point, yes there is always a danger with simulation that you are just seeing what you expect to see but it doesn’t always happen. A simulation isn’t like a mathematical proof, it doesn’t give us certainty that the model is valid but it is one more data point that can support or detract from a model.

    • In reply to #3 by Lonevoice:

      …isn’t it obvious that the resulting simulation will be similar to what we see because we have assumed both the starting point and the development model? In other words: all we’ve got is an animated version of what some people think.

      The simulation is much more than that. It does start with initial conditions put into the sim plus modeling of certain physical processes (as many as the computer can handle anyway). The initial conditions are those given by the CMB which places some pretty tight constraints on what kind of stuff the universe is made of. The real assumptions are some of the physical processes (such as the model of how gas forms an accretion disk around black holes). This simulation is not just what some people think; the material content is from myriad observations and the physics the sim follows are all the current models in astronomy (also studied individually in their own right) put together to see if it will produce a recognizable universe.

      What beats me is how anyone will ever know if any simulation about the distant past is an accurate reflection of what actually happened when no-one was there to confirm the proposed theory or say otherwise.

      Because we can see the distant past. Let’s not forget about lookback time. We can see galaxies at many different stages of evolution back to 13 billion years ago.

      This or any other simulation could be wildly off the mark but we would never know for sure – actually. So, does this simulation really show us anything?

      The physicists/astronomers who made this simulation do not just run it and look at it. They take data from it and compare it to the data from the universe. It is amazing that this simulation has enough resolution and modeled physical processes that it is the first to have galactic evolution: the sim forms spiral galaxies, elliptical galaxies and irregular galaxies. It correctly simulates the density profile of galaxies in clusters, the percentage of gas content in galaxies and the percentage of heavy elements.

      The sim (and all sims before this one) fail to accurately represent the formation of low mass galaxies. They form too early in the simulation throwing off stellar ages. This may be solved in the future by modelling stellar radiation which evidently is too computationally intensive. Radiation pressure on gases may delay star formation sufficiently to match reality.

      So this simulation gives us clues that we are on the right track when we say that the universe evolved under certain processes and it suggests to us which processes are most important to particular phenomena.

Leave a Reply