Mapping the Milky Way

We have growing evidence that many of the stars of the night sky have planets circling around them, but where did the stars themselves come from? Our galaxy, the Milky Way, is a grouping of around 100 billion stars and all the brightest stars in the night sky are part of this family. But when and where did the Milky Way's stars form? The first step to answering that question is to know how they're distributed through space. Dominic Ford spoke to Lennart Lindegren from the Lund Observatory in Sweden is working on a new space telescope that will map out the galaxy.

Lennart - Gaia is a satellite designed to survey about 1 billion stars in our galaxy and one of the main results from it will be distance determinations to many of these stars, so that we get the truly three dimensional map of our galaxy.

Dominic - The problem astronomers face is that while they can measure the projections of stars of the night sky very accurately, it's much more difficult to know how far away they are. The technique Gaia scientists are using relies on the earth's annual rotation around the sun. Just like nodding your head from side to side, this makes it possible to judge for distances to stars, by how much they appear to shift from side to side over the course of the year as Lennart's colleague, David Hobbs explains...

David - So basically, the earth just goes around the sun once every year and our satellite is somewhere around the earth of course. By looking at nearby stars, you see they're in one position in July and then if you look at them in January, you see they've shifted to another position. If you actually do that and you can do it in nice video plots, you see that it traces out a nice oval on the sky and then the angle of that oval gives you the parallax measurement which you can convert then into a distance with a simple formula.

Lennart - So, over the 5 years that Gaia will be working, the stars will be seeing to wobble back and forth by a tiny amount, depending on their distance.  The nearer the star is, the bigger is this wobble. So, by measuring this small angle, we can get the distance to the stars.

Dominic - So, these stars are nodding back and forth in the sky. I guess that motion must be very small, given how distant these stars are from us.

Lennart - Yeah, when we show videos of this wobbling of course, we exaggerate it by a factor of 100,000 or something like that and that's the reason why you need a very highly precise satellite to do these measurements.

Dominic - The technique of using the parallaxes of stars to determine their distances has a long history in Sweden and Denmark. The idea was pioneered over 400 years ago by Danish astronomer Tycho Brahe who proved that the supernova he saw in 1572 was an astronomical object rather than a weather phenomenon in the earth's atmosphere as other astronomers believed at that time.

Lennart - Yes, that is right. He used that method to prove that the new star, the stella nova which was discovered in 1572 was actually further away than the moon which was a revolutionary discovery at that time. So now, we are using this method since 200 years to measure distances to stars, which is of course, much more difficult because, since the stars are more distant than the moon for example, the parallax will become very small. When we want to measure distances to stars in the other end of the galaxy for example, these stars are very, very distant. So, the parallax becomes very small and therefore, it is difficult to measure, which is why we can't do it until now.

Dominic - So, Tycho was doing this with comets back 1572. When were we first able to measure the parallax of a star?

Lennart - The first, really successful or convincing measurement of a stellar parallax was made by Friedrich Wilhelm Bessel in 1838. He was working in Konigsberg in Germany. He measured the distance to star number 61 in the constellation Cygnus. That was a real breakthrough in the history of astronomy because astronomers had tried to measure parallax for centuries and this was the first convincing detection of the parallax.

Dominic - What's Gaia adding to that? I guess, by being in space, you haven't got the distortion of the earth's atmosphere.

Lennart - Yes, that's right. That's very important. The atmosphere has been an obstacle for accurate parallax measurement until 1989 when European Space Agency launched the Hipparcos satellite which was the first satellite designed to measure parallaxes from space, and it did that very well with an accuracy of about a thousandth of an arc second. But only for about 100,000 stars and the only stars rather close to the sun. With Gaia, we want to measure many more stars and many more distant stars, and therefore, we need much more accurate measurements.

David - Gaia is 100 to 1,000 times more accurate than Hipparcos, just because of the new instrumentation but the measurement principle is basically the same.  If you think of Hipparcos and you plot what kind of scale Hipparcos could see for a solar type star for example on top of the galaxy then you'll see that Hipparcos could only see very locally. It's a little dot on the galaxy basically. It certainly wouldn't be much bigger than just sticking your pen on a piece of paper. But if you take Gaia then you can see that the distant scales that Gaia can probe for the same kind of star is very far out. For solar-like stars, Gaia can probably see out about 8 kiloparsecs with an accuracy of 10% to 20%. That in astrometry is considered to be very good. But then for very bright stars, Gaia can see right away across the galaxy.

Dominic - We often hear about exoplanets being discovered by their gravitational pull, causing stars to wobble back and forth. I guess there are lots of other phenomena that you're having to distinguish this wobble form.

David - Yeah, of course. So, what we do is we build models of how the light should enter the telescope basically. You have to take into account a great many things. Of course, the finite speed of light for example, this is known as the Roemer correction has to be put into the time measurements even. Then you have the parallax wobble as I mentioned, but also, you have the light deflection in the solar system. So, you have to have a model of general relativity which is an extremely accurate model. It must be more accurate than the final position of Gaia. So, we have to have a microarcsecond relativity model.

Dominic - So, the issue there is that the bodies in the solar system bend light because of their gravitational fields.

David - Yeah, sure and you have to take this into account and of course, the sun for example at 90 degrees to the sun, you're still getting 4,000 microarcsecond light bending. So, it's an enormous effect. You have to take the sun's light deflection into all measurements. You typically also want to take Jupiter's light deflection into account because Jupiter is a very large body also. You also interestingly enough have to take the Earth and the Moon into account, or you should take the Earth and the Moon because the light deflection is very weak from the Earth and the Moon, but they're very close to Gaia at already 1.5 million km away. So, you should also take that into account. And of course then the other planets also have some effect. We actually use the measurements of Gaia, we take all of the measurements of Gaia together and we try to use that to test, does Gaia tell us that Einstein is right for example. And the point about Gaia is, we have so many measurements that we think we can make the most accurate test of light deflection due to general relativity and Einstein's theory and so on possible.

Dominic - Obviously, it's interesting from the point of view of natural history to make a catalogue of these distances to the stars we see in the night sky. What scientifically can we get out of that catalogue?

Lennart - Well, first of all, it can give us a detailed map of the structure of our galaxy. As we see it on the night sky, we only have a 2-dimensional image of the galaxy. So, getting the third dimension is very important for understanding the large scale structure of our Milky Way. But also, the individual distances to the stars are very important. To understand their physics, you need to know the distance to a star to translate the brightness on the sky to the real luminosities of the stars and get their physical properties. So, all kinds of astronomy will benefit from this information.

Dominic - Is that structure telling you about how our galaxy actually formed, how it's evolved, where it's come from?

Lennart - That is one of the scientific aims of Gaia to try to understand the history of our galaxy. It is thought that a big galaxy like ours is partly composed of many smaller galaxies that have been eaten up by our galaxy. They have simply fallen into our galaxy and it may be possible to identify which stars came from different in falling galaxies in the past and therefore, know a little bit of the history of how the galaxy was assembled.

Dominic - Gaia is going to launch, is it November or December this year? What's happening at the moment?

Lennart - At the moment, the satellite is going through the flight acceptance review which means that the engineers and scientists responsible for putting together the satellite, check that everything is in order. It appears that it is there, so it will go ahead for launch in November, December hopefully.

David - Gaia is ready to go today. It just has to be shipped to South America and then is ready for launch. So, it has to be flown down in two parts - the sunshield and the spacecraft goes separately because the sunshield is so big. Then there's no storage facilities in French Guiana, so it has to it has to be more or less mounted once it gets there and so on, and be ready. An Antonov Russian airplane was booked to send it down there in July and then they were told, "No, you can't go." The problem is, there was a conflict with some GPS satellites being launched. Because of that conflict, Gaia got shoved back a little bit. So, we've been pushed back by 2 months now until November/December. There's a launch window in November/December which is more or less defined by where the moon is for example. You don't launch into the moon. So, we have certain dates where we can launch and we've now been assigned this new launch date from the 17^th of November to the 5^th of December. Hopefully, it will go then because the spacecraft is sitting there and all the tests are done. It's really ready to fly. People are just doing some last minute monitoring of the spacecraft at the moment.

Dominic - And once it's up, how long until we start getting scientific data back from it.

David - Yeah, well the first thing, it has to be sent into a transfer orbit to the L2 Lagrange Point which is 1.5 million km away from the earth. As it is flying out, the sunshield will be opened. The spacecraft is cooling terminally. So, once the sunshield is open, it can actually take measurements, but those measurements would be probably quite poor to begin with because the terminal cooling of the spacecraft takes a month or so. Also, the transfer journey takes a month or so, but people will actually start taking measurements because the sooner you get something even though it's very inaccurate, you can start testing your data reduction scheme with it.

Dominic - When can we expect the first scientific results to come out from that?

Lennart - Some preliminary results will come already about 2 years after the launch, but that will not be very accurate. So, as it accumulates more data and they are processed, we will have successively more accurate results and the final results which is what all the astronomers are hoping for will come around 2021. So, that is a long time to wait.