300 done, 1500 to goPosted: October 17, 2009
[tweetmeme only_single=false service=”wp.me” source=”allinthegutter”]
Last week I was in Milan doing something I traditionally don’t really do – getting my hands dirty on data, and doing my bit to make it usable for science.
By data, in this case, I mean 100,000 spectra which were collected by the VIMOS spectrograph which is one of the instruments on the Very Large Telescope in Paranal for a new project caller VIPERS. A spectrum is simply the light of an object decomposed in a relatively large number of components. So instead of decomposing an image in, say, 3 colours, a spectrum will often have hundreds to thousands of different pixels – or colours, really – so we can see exactly how “red”, or “blue” an object is. Think of it as a rainbow – which is simply the light of the Sun decomposed in a number of colours. And yes, astronomers spend a large amount of effort effectively making rainbows out of the light of galaxies.
Spectra of objects (stars, galaxies, quasars, planets, etc) are one of the most useful observables of our Universe. Depending on whether you are a cosmologists or someone who studies galaxy evolution, you want them mainly for different reasons. Today I’m going to put my cosmologist hat on.
As we’ve covered here before, galaxies have a very rich range of properties which we can use to trace their evolution. With our hat today, none of it matters – the astonishing this is that we can learn an awful lot about the evolution of the Universe as a whole simply by studying the position of these galaxies. This is the realm of observational cosmology – the study of the birth, evolution, dynamics, composition and ultimate fate of our Universe by studying the spacial distribution of galaxies. This may be a somewhat narrow description of a huge field of modern-day Astronomy, but to first order it’s perfectly correct.
So why is the spacial distribution of galaxies so revealing? Well, let us start by considering the components of our Universe. In our current standard model we have four main components: radiation, baryonic matter, dark matter and dark energy. The first two we are very familiar with – baryonic here just refers to the matter that makes up all we see in the Universe: ourselves, the solar system, distant galaxies, far away planets, etc. The last two are more of a mystery, and excitingly also the two most important components of our Universe today. I must leave a more thorough explanation for another post, but for now let us state that there is 5 times more dark matter out there than baryonic matter, but we can’t see it. It interacts gravitationally with baryonic matter so we can detect its presence and because it is so much more abundant than baryonic matter, it’s dark matter that has the most influence in the master game of tug of war that is the evolution of our Universe – not baryonic matter. The other player is dark energy. Now, I can’t tell you what dark energy is (nobody can), but I can tell you that it behaves in the opposite way to gravity. Whereas one attracts, the other repels. Whereas one brings things closer together, the other takes them apart.
The amount of dark matter and dark energy govern the dynamics of our Universe, but we can’t see either of them directly. So we turn to what we can see: baryonic matter, via the radiation produced mainly (but not exclusively) by stars. The neat thing is that baryonic and dark matter attract each other gravitationally, so baryonic matter traces dark matter. And this is why simply measuring where galaxies sit is so insightful – it gives you a tool to track dark matter and see how its spacial distribution evolves. This in turn tells you about the interplay between dark matter and dark energy – depending on which one dominates, by how much and for how long, the evolution of the spacial distribution of dark matter (and therefore galaxies) will be different. And that we can measure! The spacial distribution (or clustering) of galaxies is one of the most promising tools to tell us about dark energy and how structure has grown in our Universe.
But before all of this can be done, we need to know the distance to each galaxy. This means measuring their redshift, by using the spectrum of each galaxy to determine how fast it is receding from us – which in turn tells us how far away it is. If the data coming out of the telescope and spectrograph are good enough, the whole process can be fully automatised. However, our position is slightly different in that due to some instrumental artifacts, the red side of the spectra is much below standard and the computerised pipeline often fails in assigning the correct redshift.
This is where I (and around 25 other people) come in. Humans are much much better at spotting mistakes than computers, so each of the 100,000 spectra is being looked at by at least two different people to make sure that the redshift is correct and that when the time comes to do science, we are doing it with reliable data. As I mentioned, this is not something I’d ever done before, so my time in Milan was used to get me trained and up to scratch with the software. I measured 300 redshifts this week, which is rather slow especially when I think that I have to do 1500 more or so in the next couple of months!
It’s a huge task but one that really needs to be done. And on the upside, I will never get bored on a train or airport ever again!