[BLANK_AUDIO] So, here we are today on a windy day in Edinburgh, next to a piece of technology from 1928, this old telescope inside the dome here. But we're going to talk about much more modern technology, the computer. Now, computers are everywhere today of course but this has been true in astronomy for many years. Computers are crucial in astronomy in at least four different ways. So let's look at those one by one. So, the first thing is that computers are intimately connected with the whole idea of electronic detectors. So here we've got a picture on my laptop - an image taken Canada-France-Hawaii telescope. It shows a deep region of the sky, lots of stars and galaxies. It's very pretty but it's not just pretty, it's made of numbers. I can bring that up here... ... a little panel that shows me. So I move over the image... the actual numbers are at each spot inside the image. So we need those numbers to do science. We want to manipulate and analyze those numbers. And that means that every electronic detector has a computer connected to it. [BLANK_AUDIO] So the next area where computers are important, is perhaps not so obvious - its about telescope control. Let's see why that's important. We pick a star, we want to look at it, and it's somewhere in the sky over here. We want to wheel our telescope around and point precisely at that star. But that's very hard to do accurately. Any mechanical structure like this telescope is going to flex as you tip it around the sky. Now the atmosphere gives us problems as well. As the light from the star comes through the atmosphere, it gets bent, it gets refracted, so that the apparent position of a star it's different from its true position by around a minute of arc. What's more, that refraction will change with the temperature and pressure of the air. So it's different from one hour to the next. So, how do you get around that problem? The traditional method is as follows. You have a small telescope on the side of the big telescope which is known as a finder. It's got a wide field of view. So, we know that our targets is going to be in there somewhere. We then take a star map like this one here. We spot the pattern of stars. We locate our target. And then having spotted our star, we tweak the telescope to bring our target into the center. So that's the traditional method. But what we can do with computers is much better because those effects - the flexing, and atmospheric refraction - we can calculate them. But we have to do this very fast. We have a computer, we can do that. So, we can make the corrections in real time and point the telescope in exactly the right place. But we can do even better than that. You remember in week two, we talked about how the atmosphere, the turbulence in the atmosphere, makes little wiggles in the light that make a star blurred. That's what seeing is. But we can use computers to correct for that motion. The way we do that, the picture I'm showing you now is a laser going up into the sky. It makes a fake star that's very bright. Then as we watch that star with our telescope and detector, we track the motion of that fake star. And then using the computer, we actually bend our optics to make an inverse motion and take away the distortions of the atmosphere. So the problem is though we have to do that calculation of how to bend the optics many times a second. So this is a process known as adaptive optics but with computers and modern electronics, it works beautifully and you can take extremely sharp pictures [BLANK_AUDIO] The third area where computers are vital, is in theoretical calculations. Now let me show you a piece of data first. The animation your looking at here is from a survey of the universe made in Australia. Its a slice through the universe - it maps out the positions of galaxies and you can see the structure is absolutely fascinating. There are filaments and bubbles, and clusters. It's what's known as the cosmic web. Now, the challenge to theorists is whether they can explain that structure in the context of the Big Bang model. And to do that, you make computer simulations. They take hundreds of millions of fake matter particles, calculate the interactions between each pair of particles and work out how the whole structure will move and evolve with time under the influence of gravity. They're very simple in principle, but very difficult in practice, because the number of calculations you have to do is enormous. So to do this you need a super computer. But the results are very interesting. The movie I'm showing you now is such a simulation made by astronomers in Germany, Durham and Edinburgh. And it does a pretty good job. [BLANK_AUDIO] So finally, we want to use computers to store and to access all the data that astronomers all over the world have accumulated over many years. So ideally, we want all that data to be live on the Internet, and to make it easy to mix and match different data sets. So, for example if I found an x-ray source in some data I have I want to look at somebody's else's infra-red sky survey and say, is my source in his survey as well, and match up that data. It's a problem we face all the time. But as well as wanting to mix and match data from different places, even individual databases can be a serious challenge to work with because they can be very big. We're talking about hundreds of millions or billions of objects in a database. And searching through those for just the object you want can be a serious challenge. This is known as the the needle in a haystack problem. An example, I am showing you here - - this is the the image in which we found a quasar at redshift seven. That's the most distant quasar ever found. To find it though, we had to look through hundreds of millions of false alarms; that was very hard work. So, that's computers and what they do for astronomy. In the third video, we're going to be looking at computer technologies and how we use them what the issues are. But before then, in the next video, Catherine is going to be looking at a classic needle in the haystack problem - how to find killer rocks. [BLANK_AUDIO]
