In the old days, astronomers had to beg for telescope time. They’d put together a proposal, convince observatories to gather data for them, crunch that data and release the results. No telescope, no results. But everything’s different now. Fleets of robotic telescopes constantly scan the skies, building up a vast database of raw data about the Universe. Anyone who wants can access the information through the Internet, download what they need to do real science. No telescope necessary. Let’s look at the development of sky surveys, and how they’re changing how astronomy gets done.
Download the transcript
Fraser Cane: In the old days astronomers had to beg for telescope time. They put together a proposal, convinced observatories to gather data, crunch the data and release the results. No telescope, no science.
But everything is different now. Fleets of robotic telescopes constantly scan the skies building up a vast database of raw data about the Universe. And anyone who wants can access the information through the internet; download what they need to do real science – no telescope necessary.
Let’s look at the development of sky surveys and how they’re changing how astronomy gets done. Is that a little over the top you think?
Dr. Pamela Gay: No, I think sky surveys really are actually changing how everything is done and that’s kind of cool.
Fraser: Why don’t you then regale us with a story of you attempting to get science done the old way? [Laughter]
Pamela: I think the best example is my doctoral dissertation. I was working on a project called theSurvey where we were following up on a radio survey actually, looking for places on the sky where we thought maybe there was a higher probability of finding galaxy clusters.
We put in for observing time at McDonald Observatory, first on the 30 inch telescope, then on the 107 inch telescope. Hundreds of nights were spent out at the observatory.
When the weather went bad we didn’t get our data, I attracted clouds, and there was much crying and sadness. We got somewhere but we didn’t get where we wanted due to clouds and forest fires and all that sort of stuff affecting the amount of time we had at the end of the day. It was sad.
Fraser: It still kind of happens now. I know you’re telling me about some projects that you’re working on where you have to scrape together telescope time. You have to convince amateurs to let them do some observations for you.
You’re distracting telescope operators so you can sneak in and you know, quickly move the telescope. [Laughter]
Pamela: There are two different ways of needing to get data. There is the “I have a question that requires significant coverage on the sky. I want to look at a whole bunch of different objects and try and prove something in a statistically significant way.”
To do that you need a ton of telescope time scattered all over the sky. The data that I take might also be good for somebody else. Not the type of stuff that leads naturally to the type of surveys that we’re going to talk about in the rest of the show.
The other type of thing you need is: “Ooh, I found a really cool object! I need a bazillion hours on this one really cool object.” That’s where you start begging people. That’s where you need the dedicated time to look at just your object.
This is why the stable of astronomical equipment needs to include both telescopes that are dedicated to doing surveys, looking at the whole sky night after night and also telescopes that are set aside for astronomers to follow up on pet projects.
Fraser: Alright, let’s then segue into the survey. Cane you give us an example of sort of what was one of the first sky surveys put together? What’s involved?
Pamela: The first really big survey that people paid attention to even today, is the Palomar Observatory sky survey. It used glass plates, two different Kodak emulsions, one sensitive to red; one sensitive to blue to look at pretty much the entire northern hemisphere of the sky all the way down to minus 30 south. It had significant coverage of the sky.
The idea was to catalogue what’s out there. The survey has been used to look for galaxy clusters. It’s been used to get statistics on how many of what different types of objects are out there.
It’s used even today where some cool something or other happens and you pull out the digital sky survey and look to see what was there before. It’s a historic record of what was where on the sky when. It’s a map of the sky and it’s a way to do big science admittedly on the 1940s technological scale.
Fraser: You’ve got a really nice telescope every night, taking a picture, moving a little bit, taking a picture and moving a little bit. It’s just slowly cataloguing every single little piece of the night sky.
I guess as we talked about before to really get good science about an object, you want to point Hubble [Laughter] at it for about a hundred hours and get every stray Photon that’s coming from it. This is the opposite, right?
This is quick and dirty. This is click, move; click move and so as you said you’re able to count up galaxy clusters. You’re able to count stars. You’re able to get a general sense of what’s out there. But you’re not able to really dig deep and see the same thing pointing at one object for a hundred hours.
Pamela: No and in fact with these old glass plate surveys, it took them years to get coverage of the entire sky. Even today we’ve moved on and today’s new version – perhaps new version is too strong a word, but our new optical survey of the sky is the Sloan digital sky survey.
It’s probing the southern galactic pole. It’s looking at a very focused region of the sky and it’s studying extremely deeply. It’s doing it in 5 colors whereas the Palomar sky survey looked at the sky originally in only 2 colors and it’s getting huge swaths of the sky every night.
But to get as deep as it does, it takes significant amounts of time and it takes years to get a really detailed survey of the sky complete.
Fraser: Let’s take a look then at Sloan. When did Sloan get operating?
Pamela: They started building the telescope back in the 1990s and it first started taking images in 2000. It’s still here in 2008, reinventing itself, coming out with new ways of doing things.
Fraser: Then what does it capture? Obviously it takes a picture of a chunk of space but I know there’s more information that it’s helping gather, right?
Pamela: The Sloan digital sky survey is using a technique called drift scanning. You might start off with an object on the right-hand of the CCD and over time it slowly drifts from the right side to the left side.
As it is drifting the digital camera is taking that right-hand most column of data and shifting it one pixel to the left and shifting it one pixel to the left. It’s shifting the recording of the data one column at a time at the same rate that the object is moving across the CCD.
By the time it has read all the way across the chip you might have many, many minutes of observations of that particular object allowing you to get extremely deep images all across the sky.
Fraser: Okay, I understand, it’s like the big CCD isn’t moving so it is able then to just pick it up, see the same object again and again.
It’s almost like it’s taking multiple photographs of the same object because it’s grabbing such a big swath of the sky at the same time.
It’s using the rotation of the Earth to move the objects in it through its field of view.
Pamela: Yeah, that’s exactly what’s happening. What’s really cool is in addition to doing this detailed imaging of the sky, the Sloan digital sky survey is also going back and doing follow-up spectroscopy.
In this case they’re actually taking plates and drilling holes in them and then aligning fiber optics onto the holes on the plate. For each hole in the plate and each fiber, they get individual spectra that allow them to get a sense of what elements are in the objects that they’re looking at. What is the red shift of some of the galaxies that they’re looking at?
These are pretty big fibers so while they’re able to capture a lot of light all at once they also aren’t so good for dealing with really crowded fields like galaxy clusters.
But for isolated objects this system is allowing us to sample a huge number of galaxies scattered all across the sky to find out where they are in red shift space. That gives us a sense of their distance and to also give us a sense of what are other things we can learn by looking at the elements.
This starts to play more of a roll when we’re saying does this thing have absorption lines? You can start to differentiate different types of active galaxies by looking at what lines exist in emission.
This is where you have an angry super massive black hole in the center of the galaxy that’s chomping on things and heating them up and if the alignment is just right, the heated up elements end up radiating emission lines that we can detect. That tells us something about what’s going on down in the core of the galaxy.
Fraser: Then how much of the sky is Sloan going to be mapping?
Pamela: Sloan digital sky survey is looking to eventually map out about 25% of the sky. That doesn’t sound like a lot, but at the level of detail they’re getting this is actually a really powerful survey.
Already they’ve allowed us to learn new things about the structure of our own galaxy that we’ve never even guessed at. Galaxies are the cool part for me. In addition to looking at all the galaxies, it’s also mapping out stars in the halo of the Milky Way.
By looking at the colors of the stars we’re able to get a sense of how far away they are. So you look at the color and brightness and basically by taking a color magnitude diagram of what would a population of stars look like at what distance.
You can sample through the sky to find out where are the populations of stars that are 60 kilo parsecs away. Where are they that are 70 kilo parsecs away? We can start to see streaks, tidal tails of shredded dwarf galaxies out in the halo of the Milky Way by probing through using the Sloan digital sky survey data to find what we call co-moving populations of stars.
Fraser: I guess this is where the real power is because in the olden days you would take your telescope, look at a region of sky, and make measurements of whatever you wanted to do.
But in this situation Sloan is gathering images of everything that’s in that 25% of the sky and furthermore they’re writing down numbers. They’re saying: “there’s a star here at this location, there’s a galaxy at that location. There’s a quasar here at this location.” And then they’re figuring out what the elements are that are in those objects. You can then do searches.
This is where it becomes a database issue. You can say: “find me every star that has this level of metalicity, or find me all of the stars that are within this range in this region of the sky.” The data mining starts and perhaps suddenly you look at that line where you mapped out all the stars in the survey and they happen to be lined up in a very interesting line. That must be a tidal tail.
So there are whole new kinds of discoveries being made that could never be made before because you just didn’t have the raw data about the entire sky that you could just then mine and ask questions.
I know there’s some amazing work done for dark matter, quasars, as you said cataloguing active galaxies. Isn’t Galaxy Zoo using the Sloan digital sky survey?
Pamela: Right but before we jump on to galaxies I just want to make it really clear that the spectra that they’re getting with the Sloan digital sky survey really isn’t good enough to start getting at detailed metalicities of stars. It is rough spectra but it’s enough to get us what are their velocities and to get us broad information on them.
That’s what’s cool about the Sloan digital sky survey. It gives us a broad understanding of where things are, of how they’re moving, of what different types of things are out there that we can follow up on later.
It is allowing us to find more white dwarfs than we ever thought we could find some other way. It is allowing us to find quasars. It is in fact allowing us to determine the distribution of galaxies of different shapes and sizes and orientations on the sky. This is where Galaxy Zoo comes in.
Fraser: People are finding asteroids and kuiper objects that are sort of in the pictures.
Pamela: They’re starting to find really rare objects. One of the really cool things that came out of the Galaxy Zoo 1 project, a project that was originated by folks over at Oxford, my collaborator Chris Lintott, Kevin who is now at Yale, a whole bunch of different folks, were sitting around.
Kevin was given the task to go look at 50,000 galaxies and tell us where their distribution is in terms of are they spirals, are the elliptical? How are they oriented? After doing 50,000 objects he really didn’t want to do any more.
So in a pub, the idea was originated to get the public who likes looking at galaxies to look at almost a million objects from the Sloan digital sky survey and catalogue how they’re oriented and what their shape is. Basically are they clockwise, counterclockwise or edge-on spirals or are they ellipticals or are they mergers? They did this and the results of the public looking at all of these objects were just as accurate as getting a small number of professionals to look at a much smaller sample of objects because you can’t make professionals look at too many of them before we go crazy.
You ask 170,000 people to look at things and of course 170,000 people can look at a lot more objects than 5 or 6 professionals.
Fraser: Right and you’re saying that they turned up some amazing things.
Pamela: One of the really cool things that they were able to do is build up a catalogue of galaxies that are overlapping on the sky. These aren’t galaxies that are merging but rather two galaxies that are superimposed on one another’s line of sight, we say. One is nearby, one is further away.
Gravitationally they really don’t care about each other very much, but the background galaxy can act like a spotlight going through the dust lanes of the foreground galaxy allowing us to make out the details in structure that we might not otherwise be able to see.
Fraser: I guess it’s almost impossible for a computer to seek out those kinds of objects because it can barely tell that there’s a galaxy there at all. This is something that a human being is great at.
Pamela: Computers are generally pretty good about determining what is a star and what is not a star. Beyond going not a star, computers tend to get kind of confused.
At a certain level you can start to program them to look for things that are S-shaped or Z-shaped so we have clockwise and counterclockwise shaped galaxies.
We can program them to look for things that have a radial profile, things that are basically fuzzy blobs like elliptical galaxies. You can’t program a computer to determine if this thing looks like nothing anyone has ever seen before. They’d be showing that up every time an asteroid happens to pass in front the galaxy.
You can train human beings to determine what an asteroid looks like when it is passing through an image; this is what a satellite looks like when it’s passing through an image. This is a really weird nebula; this is reflected light inside the telescope.
We can take human beings and train them to do things that we can’t train a computer to do. Human beings will naturally note “this is cool and unusual” and bring it to the attention of other people. That’s one of the wonderful things that came out of the original Galaxy Zoo project. What’s even cooler is there is now a second generation, Galaxy Zoo 2 that’s been launched.
If you go to galaxyzoo.org there is now a link off of it to Galaxy Zoo 2 but you’re going to have to take a survey if you participated in Z1 so that some of the folks working with Galaxy Zoo, which in this case includes me, so please take the survey, can find out a little bit about why is it that you love using Galaxy Zoo. I know why I love using it; I want to know why you love using it.
Fraser: What are some other surveys that are happening?
Pamela: Surveys aren’t restricted to just being optical telescopes. There are lots of other surveys out there. The Very Large Array out in New Mexico is working on a survey called ‘First’ and it’s looking to do both the northern and southern galactic poles.
It’s a survey that’s looking at 21 centimeter continuing radiation. This is the type of light you get from disks of galaxies from blobs of gas. It allows you to see star-forming regions. It also allows you to see jets off of radio galaxies.
The first survey has about 5 arc second resolution which is what you get on a really bad day with an optical telescope. It’s about 5 times what you’d get from a reasonable sight with an amateur telescope. It’s going pretty faint and it’s looking to collect data that will allow us to figure out where are all the galaxies that are actively forming stars.
Where are the most distant radio galaxies – galaxies that are actively feeding black holes in their center and are thus giving off radio emission? This is an ongoing survey that’s constantly working to increase its area and it’s working in radio. It’s just another way of looking at the Universe.
Fraser: So, same deal, right? Gather as much raw data as you can, catalogue it as best you can and then make that information available to the scientific community and I mean the general public – it’s all on the internet – to look for whatever they want in the data as it stands which is just amazing.
Pamela: And we’re working to try and cover as much of the sky as we can and as many different colors as we can. There’s 2MASS out there working in micron radiation that used two 1.3 meter telescopes in Arizona and Chile to look at the sky.
There have been infrared satellites and x-ray satellites that have also worked to cover the sky. We also look at data that was perhaps taken for other purposes as another source of perhaps serendipitous observations.
There’s a telescope called the near-Earth asteroid telescope, neat. It is primarily out there trying to make sure nothing hits the planet Earth. It’s a good goal, but as it is out there surveying for this very specific purpose, it’s also picking up supernova. It’s also picking up variable stars. It’s picking up lots of other stuff that just happens to be in the background of the same fields it is looking for asteroids in. We can use that data as well.
There are very specific surveys looking at very small regions of the sky but because they’re building up data over years, we’re able to learn interesting things. There were two projects, the MACHO project and the OGLE project that looked at the magellanic clouds specifically looking for gravitational lensing events. This is where a nearby object, in this case nearby being on the outskirts of the Milky Way Galaxy, passes in front of the background object, something in one of the magellanic clouds and causes it to brighten through gravitational micro-lensing.
In the process of looking for these events, they’ve also done things like find light echoes from supernova moving through the interstellar medium. So there is a supernova hundreds of years ago, thousands of years ago and the flash of light from that supernova is forming an expanding shell of light. As that light passes through the gas and dust between the stars, it temporarily illuminates whatever section that it happens to be in.
By looking in the same direction for year after year after year, we can watch these shells of light move and then track them backwards and figure out where they originated.
This is one of the neat ways that we’re starting to figure out what was Kepler’s supernova actually like? What was the supernova in Cassiopeia actually like? We’re tracing back the light echoes to learn more about events that we weren’t around to see.
Fraser: This I think is a realm of science that has no limit. You can just imagine bigger telescopes, more telescopes gathering more of the data that are looking deeper. Are there sort of dream projects in the works to do really enormous surveys?
Pamela: What’s really cool is they’re not even dream projects, they’re actual projects. There are two really cool ones coming up, Pan-STAARS and LSST (Large Synoptic Survey Telescope).
These two different many meter telescopes are focused on trying to find things that are going to hit the planet Earth. Protecting the planet Earth is a good way to get money to build telescopes. Along the way while they’re out there taking snapshot after snapshot after snapshot of the sky they’re also going to be turning up supernovas. They’re going to be turning up variable stars. They’re going to be taking image after image of the same place on the sky.
Those images can be added together to get some of the deepest images we’ll have ever achieved of distant galaxies. It’s all a matter of adding up the data over time. With these two projects we’re going to have these huge telescopes taking in pretty much everything that’s visible every single night.
That data can get added up to allow deep imaging or it can get used together to get time sequence imaging to see what are all the transient events that we’ve been missing all these years.
Fraser: I think just to be clear all this data is available on the internet, right? If you know where to look, you can pull it down and crunch it.
Pamela: Different surveys have different proprietary periods. This is a certain amount of time where the people who invested the intellectual resources and the monetary resources to build a given instrument (telescope) or take a certain survey, get to have all the data to themselves. At the end of these proprietary periods all of the data becomes publicly accessible.
One of the greatest ways to go out there and access a lot of this data is through a portal called Sky View. Just open up Google and do Sky View Virtual Observatory. It will give you a form to fill out that will allow you to get radio data, x-ray data, and optical data, all of the same field on the sky.
It will allow you to map different objects onto these images to see where they happen to line up. It’s a great resource for both the professional astronomer and for people who are just trying to get a different wavelengths perspective of the Universe.
Everything is out there. Sometimes you just have to wait 6 months to a year to get your hands on it.
Fraser: I think the big need is for people in the computer industry, people who understand how to make a database sing and to be able to pull in that data and help crunch and help answer some of those basic questions.
But I think the reality is that anybody who wants, assuming they have the skills, can go onto those surveys, download the information and discover brand new objects, discover asteroids that have never been seen before. It’s all there; there are mysteries inside that data. All you have to do is go looking.
Pamela: And I can’t restate what you’re saying with enough emphasis. It is everyday people who are going out and discovering new things in some of these surveys. I’ve been at American Association of Variable Star Observers meetings where amateur astronomers have stood up and given talks on how they’ve gone through this database or that database pulling up variables that no one had known about before just because they knew how to do all the nice equal queries effectively.
With the Galaxy Zoo project there have been just everyday members who are participating and noticing things that are new that are making cool discoveries where go check out the forums. Look up the peas in theThese are a new class of galaxies that were discovered by regular people who asked: “What are all these little green compact galaxies?”
There’s new stuff out there just waiting to be discovered. You might be the person to make the next cool discovery.
Fraser: And if you do let us know. We’d love to hear it. Well thanks a lot Pamela and we’ll talk to you next week.
Pamela: One more thing before we quit. Hopefully next summer I will both have my voice back and I will have the opportunity, perhaps with even you, to go out and see a really cool eclipse out in the Pacific Ocean. I’m going to be on the eclipseofthecentury.com tour. There are still seats available.
Fraser: That would be fun. I’d love to see an eclipse. I’ve never seen a total solar eclipse.