Monday, 23 April 2012

A truly super computer

A short post, but the greatest computer ever built, the ZX Spectrum is 30 years old. Check it out!

I've written about this machine before, and it's influence on me, but one can only truly understand its greatness in song.
Yes, the graphics were not as spiffy as a PS3, and yes, games took forever to load (and there was the terror of tape loading errors), but you could program it in a real language (which let you play around with raw memory). For a certain generation, this song will bring a tear to your eye.

And if you don't know who the disembodied head is singing later on in this song, shame on you, shame!!!    
Through him, and the spectrum, I feel I had a distinct advantage entering university and becoming an astrophysicist. Thank you, Clive.

Sunday, 22 April 2012

Drawing the line

There has been some noise out there in the matrix about a recent book by Lawrence Krauss called "A Universe from Nothing".


It has been seized upon by by the atheist community (and to lay my cards on the table, I am an atheist) as another nail in the coffin for the God Delusion.

It has gone so far that Richard Dawkins (husband of the wonderful Lalla Ward, one of the Doctor's greatest companions) has said in the forward
"This could potentially be the most important scientific book with implications for supernaturalism since Darwin."
and
“Even the last remaining trump card of the theologian, ‘Why is there something rather than nothing?,’ shrivels up before your eyes as you read these pages. If ‘On the Origin of Species’ was biology’s deadliest blow to super­naturalism, we may come to see ‘A Universe From Nothing’ as the equivalent from cosmology. The title means exactly what it says. And what it says is ­devastating.”
I haven't read the book (but will once I have gotten Violent London and Dr Euler out of the way), but, sorry, this madness has to stop. We need to draw a firm line in the sand on what is science, and what is not (and as I hope to show, a distinct line isn't even possible!).

Where to start. I think the starting place is a rather robust discussion I had with colleagues when I was recently in France. Some may have called it an argument, others might suggest that I was vigorously lecturing the others in the room, but robust it was. The point of discussion, in summary, was the dodgyness of modern cosmology.

The discussion ranged far and wide, and I was attempting to defend cosmology. The problem was that word cosmology was used to encompass many things, from the redshift of galaxies, the temperature of the CMB and baryon wiggles, through dark matter, dark energy and quintessence, and into mutiverses, brane cosmologies and freaky cosmologies.

The problem was that using a single word, cosmology, to encompass all of these things, was then used to paint the entire topic with a single brush, suggesting that because papers on multiverses are often daffy in the extreme, and that there is no physical evidence for colliding branes, that the entire topic should be dismissed as pseudoscience and thrown in the bin.

I feel my blood pressure rising again, so here is a calming picture.

Like a lot of things, cosmology is not a single "thing" but a whole collection of observations, theoretical models, guesswork, and darn-right fairy tales.

There are some things we are pretty certain of. Galaxies out there are redshifted, and that redshift increases with distance. I have measured this myself for a handful of galaxies, and I know people who have collectively measured the redshifts for millions of galaxies. Either there is something completely screwed with the laws of physics, or we are pretty certain of this. I would bet my house that galaxies are redshifted - it's an observation thing.

Now comes the tricky bit. Why are they redshifted. Well, the only physical theory we really have that lets us calculate properties of the universe at large is general relativity (and this, this exceptionally vital point, is often lost in the media haze surrounding cosmology), and that actually predicts what we see. Galaxies should be moving away from us. But only if the universe "expands", such that the distance between objects get larger with time.  I am a big fan of relativity and so I would put my car on this interpretation of what we see being correct.

But there is a problem. To get the rate of expansion with distance that we see, we need a lot more matter than we can actually observe (i.e. all the stars and gas out there), and so we need a dark matter component. Worse than that, we need to add an even more mysterious dark energy component to get the universe to accelerate.    

Sometimes in the media, dark matter and dark energy sound like something a bunch of tipsy cosmologists dreamt up of in the pub just to annoy everyone else. But they are not. In relativity they are precisely defined in terms of their large scale properties. If you don't have them in the equations, then it just doesn't work.

Yes, general relativity may be wrong, and so we could be being fooled, but I'm willing to bet a few hundred dollars this is not the case, that general relativity is a pretty good description of the expansion of the universe, and so we need a dark sector to account for what we see.

While general relativity tells you about the bulk properties of the dark sector, it does not (and, more importantly, cannot) tell you what dark matter and dark energy actually are. This starts to mean that bets are off and speculation begins.

As long as it is dark, and behaves like matter, it is a dark matter candidate.

As long as it is dark, and causes the universe to accelerate (like a cosmological constant), it is a dark energy candidate.

For any particular candidate, I might be willing to bet a dollar here or there, but really, I am not convinced of a lot of ideas. Actually, quite a lot of them belong in science fiction.

And then we hit the biggy. If we run the universe backwards it has a time zero, a big bang where the distance between any pair of points is zero. The birth of the universe.          

BOOOOOMMMM!!!!!

But where did it come from? Surely this is where science meets religion? If science can show that the universe arose in a purely physical process, then that puts an end to the "God did it" excuse?

Yes, it does, but we cannot answer that question. And to understand why is easy. Here's the history of the universe.
where closer to today is at the bottom. As we work our way upwards, the two key aspects of physics are at work, with general relativity controlling the expansion of the universe, while quantum mechanics does all of the cooking of particles, atoms and molecules. And it works wonderfully, until that annoying time at the top. The Planck Time!

There is a lot of fluff written about quantities with the word Planck in front of them, but (as yet) they are really a sign where something has gone wrong. We know that at and before the Planck time our knowledge of the laws of physics fails. At this earliest epoch of the universe, both quantum mechanics and general relativity influence all of the processing going on, and we need them to work together, and what we know is that they simply don't (I'll write more on this at another time).

When we reach this point, there is no way we can use the laws of physics to wind back the clock and ask what happened before. If there was *only* general relativity, we know that the distance between objects goes to zero and we have the Big Bang, but with quantum mechanics in the mix, all bets are off and, and I'll put this in bold, until we have a mathematical theory of gravity and the other forces that work together in the immense temperatures and density of the early universe, we simply cannot see before this point.

Here is a picture I made a number of years ago to illustrate this (this is, in fact, an updated version by the very smart, and Bayesian-master, Brendon Brewer)



The solid black line is the "general relativity only" cosmology, the purple is the Planck time (not to scale) and the green is what really might have happened to the universe. Basically, just about anything.

And so we reach the question of "what happened before the Big Bang?". Honestly, we don't know and we reach the realm of speculative cosmology. Brane collisions, vacuum fluctuations, self-reproducing universes, the myriad ideas that are out there. But again, while they may have mathematics behind them, they are still little more than speculation. The best they can do, in terms of observational evidence, is match the standard cosmological models in terms of predictions.

This is important. There is not a shred of evidence to distinguish all of the multitude of speculative "before the Big Bang" models from the standard general relativistic picture. We know that our laws of physics break down in the early universe, but, who knows, the general relativity picture may be correct and the universe was really born out of nothing, a real nothing. Until we get our physics to work back through the Planck time, we just don't know.

This brings us back to Krauss's book (and again, I have not read it, but will and comment). While there are a number ideas presented in there for pre-Big Bang universes, as of yet, they may all be completely and utterly wrong and what really happened may be something completely different.

There is nothing wrong with presenting such ideas, as they may be the starting point for really understanding the universe. But I think it is essential for us to be honest that what we have now are nothing really than a collection of ideas, speculations and, in some cases, fairy tales, and to suggest that we have really started to address the question of what happened before the Big Bang, is somewhat dishonest. The question is more open than evolution, even in the time of Darwin.

I realize I have written a lot here, so I'm going to take a breather, and take Mini-Me #1 to rugby. More soon.

Saturday, 14 April 2012

Falling into a Black Hole

For someone who works in relativity, you'd think I'd know where the time goes! Things have been quite busy with grants, Easter and packing up my office as the building I am in is to be demolished. We're going to be off campus for a couple of years, in the lovely suburb of Redfern; at least we'll be closer to the  train station :)

This has meant I've not had a lot of time to get a lot of research done, so I thought I would post a little lesson on relativity, namely what happens when you fall into a black hole. Not the painful bit where you get torn limb from limb, but what you see.

I've worked on this in the past, but this post was sparked by this article, especially this statement.

"First of all, you approach the speed of light as you fall into the black hole. So the faster you 
move through space, the slower you move through time," he said. "Furthermore, as you fall, there are things that have been falling in front of you that have experienced an even greater 'time dilation' than you have. So if you're able to look forward toward the black hole, you see every object that has fallen into it in the past. And then if you look backwards, you'll be able to see everything that will ever fall into the black hole behind you. 
"So the upshot is, you'll get to see the entire history of that spot in the universe simultaneously," he said, "from the Big Bang all the way into the distant future."

This is, to put it politely, simply wrong, and is a common misconception in relativity.

Firstly, a little history (and remember that I am not a historian of science). A black hole is a completely collapsed object, where gravity has overcome all other forces and the all the mass is compressed into a single point (the singularity - que stupid mystical wooooooo music).

The first person to write down the metric that describes the space-time of a black hole was Karl Schwarzschild in 1916. He realised that something weird occurred at the origin (the singularity - music again), but also at the "Schwarzschild Radius"


(pinched from wikipedia). The metric "blows up" at this point, and there appears to be another singularity. Unlike the singularity at the origin, the curvature of space-time at the Schwarzschild radius is well behaved. But it looked like this singularity stopped objects, both massive and photons, entering the black hole.

Hopefully this explains it


This shows radius along the x-axis, and time on the y-axis, and the solid lines are the paths of light rays. The central singularity is at r=0, and r=2m is the Schwarzschild radius. Notice that light rays are not moving at 45 degrees in this picture (which is what you get when you consider the flat space-time of special relativity). The key point is that light rays outside of r=2m stay outside, while those inside stay inside.

This confuddled many people, including Einstein, who concluded that black holes cannot form, as nothing can get inside them.

The problem was solved when it was realised that the problem at r=2m is a coordinate singularity, which is something like asking the question "What is north of the north pole?". With a change of coordinates, to Eddington-Finkelstein Coordinates, the situation at r=2m gets smoother out.

Now, what do you get for light rays?


There are some issues with this representation, as the "time" on the y-axis is more complicated than before. But the key point is that light rays from the outside can now cross to the inside. The light rays inside r=2m are stuck in there, showing that the Schwarzschild radius is an event horizon, a one-way membrane through which things can flow.

The dotted path shows what happens to an infalling observer. There are a couple of things you can see immediately from this picture;


  • To see the entire future of the universe, the infalling observer would need to receive all of the inward traveling photons. Clearly, they hit r=0 (and certain death) before the top two photons reach them. So they only see a finite time into the future, not the entire future history of the universe.
  • They can receive light from things that have fallen into the black hole before us, but we can't see thing outside from times before we fell in (well, no more than just seeing things far away in the past due to light travel times.
If you really want to understand, take a look at Andrew Hamilton's excellent webpage.

And if you're really looking to understand, you can have a read of my not-so-excellent paper


It has long been known that once you cross the event horizon of a black hole, your destiny lies at the central singularity, irrespective of what you do. Furthermore, your demise will occur in a finite amount of proper time. In this paper, the use of rockets in extending the amount of time before the collision with the central singularity is examined. In general, the use of such rockets can increase your remaining time, but only up to a maximum value; this is at odds with the ``more you struggle, the less time you have'' statement that is sometimes discussed in relation to black holes. The derived equations are simple to solve numerically and the framework can be employed as a teaching tool for general relativity.

Monday, 2 April 2012

How science is done

Hmmmm - I originally wrote this while twiddling my thumbs at Changi Airport in Singapore for a few hours, but the "one the road" blog software I was using seems to have deleted it (perhaps it had a read and didn't like the post :)

So, I was returning from a very busy and fruitful week in France. While there, I got to thinking about just how we do science. If you read textbooks (or, if there is nothing left on the shelf worth reading, philosophy) then you get told science is testing ideas and models with experimentation, and then reworking your ideas in light of new evidence. The problem is that this viewpoint, however, is that it is too clean, to clinical, too robotic.

But let's start with the environment, the beautiful city of Strasbourg, in Alsace, France. It is home to the very impressive Strasbourg Cathedral, which, between the years of 1647 to 1874 was the tallest building in the world.
The keen-eyed amongst you will recognise this view from the recent movie, Sherlock Holmes: A Game of Shadows. While the architecture is pretty cool, my week was dedicated to science.

My visit was to the Strasbourg Observatory to work with my close collaborator, Rodrigo Ibata. So, how does this gel with the title of the post? Well, science is all about communication, and while scientists generally use email, nothing beats getting together with fellow scientists. While, because science is actually messy. Ideas are thrown out, shot down, picked up, compared to the literature, laughed at and even advanced.

As an example, here we are using the latest scientific techniques to consider trends in data.
This is called "sticking a bit of paper over a computer screen" and it works very well.

Part of the visit was spent outside of Strasbourg, out in the countryside, where we could focus even more on the research. The area was wonderfully gorgeous
and we spent a few days immersed in the discussion of science. It did not stop, from the rising to the setting of the Sun. Sure, discussions roamed from the nature of science, the rickerty nature of cosmology beyond the concordance model, the academic job market, the European economy, but the focus was science. Here's a typical scene:
Tap tap tap on the laptop, then discussions and arguments, followed by discussions and ideas (and multiple cups of tea and coffee). We managed to stop for food now and again in the warm air of spring
but while the keyboard tapping stopped, the discussions of science, techniques, mathematics etc continued.

So, it was messy, no real structure, just a general direction of conversations. And the result of the week? Extremely successful. We finished the week with a map of more than a dozen papers we need to write in the next year, most of them the most vaulted of papers - "high impact" papers, and a couple very high impact. We have ideas of new student projects, and grants to write, and directions of research to test, and data to obtain.

It was nothing like the "textbook" version of science. But it was great, and and great science was done. Nothing beats the messy interaction of real science.