Return to Video

How to take a picture of a black hole

  • 0:01 - 0:03
    In the movie "Interstellar,"
  • 0:03 - 0:07
    we get an up-close look
    at a supermassive black hole.
  • 0:07 - 0:09
    Set against a backdrop of bright gas,
  • 0:09 - 0:11
    the black hole's massive
    gravitational pull
  • 0:11 - 0:12
    bends light into a ring.
  • 0:12 - 0:15
    However, this isn't a real photograph,
  • 0:15 - 0:16
    but a computer graphic rendering --
  • 0:16 - 0:20
    an artistic interpretation
    of what a black hole might look like.
  • 0:20 - 0:22
    A hundred years ago,
  • 0:22 - 0:25
    Albert Einstein first published
    his theory of general relativity.
  • 0:25 - 0:27
    In the years since then,
  • 0:27 - 0:30
    scientists have provided
    a lot of evidence in support of it.
  • 0:30 - 0:33
    But one thing predicted
    from this theory, black holes,
  • 0:33 - 0:35
    still have not been directly observed.
  • 0:35 - 0:38
    Although we have some idea
    as to what a black hole might look like,
  • 0:38 - 0:41
    we've never actually taken
    a picture of one before.
  • 0:41 - 0:45
    However, you might be surprised to know
    that that may soon change.
  • 0:45 - 0:50
    We may be seeing our first picture
    of a black hole in the next couple years.
  • 0:50 - 0:54
    Getting this first picture will come down
    to an international team of scientists,
  • 0:54 - 0:55
    an Earth-sized telescope
  • 0:55 - 0:58
    and an algorithm that puts together
    the final picture.
  • 0:58 - 1:02
    Although I won't be able to show you
    a real picture of a black hole today,
  • 1:02 - 1:05
    I'd like to give you a brief glimpse
    into the effort involved
  • 1:05 - 1:06
    in getting that first picture.
  • 1:07 - 1:09
    My name is Katie Bouman,
  • 1:09 - 1:12
    and I'm a PhD student at MIT.
  • 1:12 - 1:14
    I do research in a computer science lab
  • 1:14 - 1:17
    that works on making computers
    see through images and video.
  • 1:17 - 1:19
    But although I'm not an astronomer,
  • 1:19 - 1:20
    today I'd like to show you
  • 1:20 - 1:23
    how I've been able to contribute
    to this exciting project.
  • 1:23 - 1:26
    If you go out past
    the bright city lights tonight,
  • 1:26 - 1:29
    you may just be lucky enough
    to see a stunning view
  • 1:29 - 1:30
    of the Milky Way Galaxy.
  • 1:30 - 1:33
    And if you could zoom past
    millions of stars,
  • 1:33 - 1:36
    26,000 light-years toward the heart
    of the spiraling Milky Way,
  • 1:36 - 1:40
    we'd eventually reach
    a cluster of stars right at the center.
  • 1:40 - 1:43
    Peering past all the galactic dust
    with infrared telescopes,
  • 1:43 - 1:47
    astronomers have watched these stars
    for over 16 years.
  • 1:47 - 1:51
    But it's what they don't see
    that is the most spectacular.
  • 1:51 - 1:54
    These stars seem to orbit
    an invisible object.
  • 1:54 - 1:56
    By tracking the paths of these stars,
  • 1:56 - 1:57
    astronomers have concluded
  • 1:57 - 2:01
    that the only thing small and heavy
    enough to cause this motion
  • 2:01 - 2:03
    is a supermassive black hole --
  • 2:03 - 2:07
    an object so dense that it sucks up
    anything that ventures too close --
  • 2:07 - 2:08
    even light.
  • 2:08 - 2:11
    But what happens if we were
    to zoom in even further?
  • 2:11 - 2:16
    Is it possible to see something
    that, by definition, is impossible to see?
  • 2:17 - 2:20
    Well, it turns out that if we were
    to zoom in at radio wavelengths,
  • 2:20 - 2:22
    we'd expect to see a ring of light
  • 2:22 - 2:24
    caused by the gravitational
    lensing of hot plasma
  • 2:24 - 2:26
    zipping around the black hole.
  • 2:26 - 2:27
    In other words,
  • 2:27 - 2:30
    the black hole casts a shadow
    on this backdrop of bright material,
  • 2:30 - 2:32
    carving out a sphere of darkness.
  • 2:32 - 2:36
    This bright ring reveals
    the black hole's event horizon,
  • 2:36 - 2:38
    where the gravitational pull
    becomes so great
  • 2:38 - 2:40
    that not even light can escape.
  • 2:40 - 2:43
    Einstein's equations predict
    the size and shape of this ring,
  • 2:43 - 2:46
    so taking a picture of it
    wouldn't only be really cool,
  • 2:46 - 2:48
    it would also help to verify
    that these equations hold
  • 2:48 - 2:51
    in the extreme conditions
    around the black hole.
  • 2:51 - 2:53
    However, this black hole
    is so far away from us,
  • 2:53 - 2:57
    that from Earth, this ring appears
    incredibly small --
  • 2:57 - 3:00
    the same size to us as an orange
    on the surface of the moon.
  • 3:01 - 3:04
    That makes taking a picture of it
    extremely difficult.
  • 3:05 - 3:06
    Why is that?
  • 3:07 - 3:10
    Well, it all comes down
    to a simple equation.
  • 3:10 - 3:12
    Due to a phenomenon called diffraction,
  • 3:12 - 3:14
    there are fundamental limits
  • 3:14 - 3:16
    to the smallest objects
    that we can possibly see.
  • 3:17 - 3:20
    This governing equation says
    that in order to see smaller and smaller,
  • 3:20 - 3:23
    we need to make our telescope
    bigger and bigger.
  • 3:23 - 3:26
    But even with the most powerful
    optical telescopes here on Earth,
  • 3:26 - 3:29
    we can't even get close
    to the resolution necessary
  • 3:29 - 3:31
    to image on the surface of the moon.
  • 3:31 - 3:34
    In fact, here I show one of the highest
    resolution images ever taken
  • 3:34 - 3:36
    of the moon from Earth.
  • 3:36 - 3:38
    It contains roughly 13,000 pixels,
  • 3:38 - 3:43
    and yet each pixel would contain
    over 1.5 million oranges.
  • 3:43 - 3:45
    So how big of a telescope do we need
  • 3:45 - 3:48
    in order to see an orange
    on the surface of the moon
  • 3:48 - 3:50
    and, by extension, our black hole?
  • 3:50 - 3:53
    Well, it turns out
    that by crunching the numbers,
  • 3:53 - 3:55
    you can easily calculate
    that we would need a telescope
  • 3:55 - 3:57
    the size of the entire Earth.
  • 3:57 - 3:58
    (Laughter)
  • 3:58 - 4:00
    If we could build
    this Earth-sized telescope,
  • 4:00 - 4:03
    we could just start to make out
    that distinctive ring of light
  • 4:03 - 4:05
    indicative of the black
    hole's event horizon.
  • 4:05 - 4:08
    Although this picture wouldn't contain
    all the detail we see
  • 4:08 - 4:10
    in computer graphic renderings,
  • 4:10 - 4:12
    it would allow us to safely get
    our first glimpse
  • 4:12 - 4:14
    of the immediate environment
    around a black hole.
  • 4:14 - 4:16
    However, as you can imagine,
  • 4:16 - 4:20
    building a single-dish telescope
    the size of the Earth is impossible.
  • 4:20 - 4:22
    But in the famous words of Mick Jagger,
  • 4:22 - 4:23
    "You can't always get what you want,
  • 4:23 - 4:26
    but if you try sometimes,
    you just might find
  • 4:26 - 4:27
    you get what you need."
  • 4:27 - 4:29
    And by connecting telescopes
    from around the world,
  • 4:29 - 4:33
    an international collaboration
    called the Event Horizon Telescope
  • 4:33 - 4:36
    is creating a computational telescope
    the size of the Earth,
  • 4:36 - 4:38
    capable of resolving structure
  • 4:38 - 4:40
    on the scale of a black
    hole's event horizon.
  • 4:40 - 4:43
    This network of telescopes is scheduled
    to take its very first picture
  • 4:43 - 4:45
    of a black hole next year.
  • 4:45 - 4:49
    Each telescope in the worldwide
    network works together.
  • 4:49 - 4:51
    Linked through the precise timing
    of atomic clocks,
  • 4:51 - 4:54
    teams of researchers at each
    of the sites freeze light
  • 4:54 - 4:57
    by collecting thousands
    of terabytes of data.
  • 4:57 - 5:02
    This data is then processed in a lab
    right here in Massachusetts.
  • 5:02 - 5:04
    So how does this even work?
  • 5:04 - 5:07
    Remember if we want to see the black hole
    in the center of our galaxy,
  • 5:07 - 5:10
    we need to build this impossibly large
    Earth-sized telescope?
  • 5:10 - 5:12
    For just a second,
    let's pretend we could build
  • 5:12 - 5:14
    a telescope the size of the Earth.
  • 5:14 - 5:17
    This would be a little bit
    like turning the Earth
  • 5:17 - 5:19
    into a giant spinning disco ball.
  • 5:19 - 5:21
    Each individual mirror would collect light
  • 5:21 - 5:23
    that we could then combine
    together to make a picture.
  • 5:23 - 5:26
    However, now let's say
    we remove most of those mirrors
  • 5:26 - 5:28
    so only a few remained.
  • 5:28 - 5:31
    We could still try to combine
    this information together,
  • 5:31 - 5:33
    but now there are a lot of holes.
  • 5:33 - 5:37
    These remaining mirrors represent
    the locations where we have telescopes.
  • 5:37 - 5:42
    This is an incredibly small number
    of measurements to make a picture from.
  • 5:42 - 5:45
    But although we only collect light
    at a few telescope locations,
  • 5:45 - 5:49
    as the Earth rotates, we get to see
    other new measurements.
  • 5:49 - 5:53
    In other words, as the disco ball spins,
    those mirrors change locations
  • 5:53 - 5:56
    and we get to observe
    different parts of the image.
  • 5:56 - 6:00
    The imaging algorithms we develop
    fill in the missing gaps of the disco ball
  • 6:00 - 6:03
    in order to reconstruct
    the underlying black hole image.
  • 6:03 - 6:05
    If we had telescopes located
    everywhere on the globe --
  • 6:05 - 6:07
    in other words, the entire disco ball --
  • 6:07 - 6:09
    this would be trivial.
  • 6:09 - 6:12
    However, we only see a few samples,
    and for that reason,
  • 6:12 - 6:14
    there are an infinite number
    of possible images
  • 6:14 - 6:17
    that are perfectly consistent
    with our telescope measurements.
  • 6:17 - 6:20
    However, not all images are created equal.
  • 6:21 - 6:25
    Some of those images look more like
    what we think of as images than others.
  • 6:25 - 6:29
    And so, my role in helping to take
    the first image of a black hole
  • 6:29 - 6:32
    is to design algorithms that find
    the most reasonable image
  • 6:32 - 6:34
    that also fits the telescope measurements.
  • 6:35 - 6:39
    Just as a forensic sketch artist
    uses limited descriptions
  • 6:39 - 6:42
    to piece together a picture using
    their knowledge of face structure,
  • 6:42 - 6:46
    the imaging algorithms I develop
    use our limited telescope data
  • 6:46 - 6:50
    to guide us to a picture that also
    looks like stuff in our universe.
  • 6:50 - 6:54
    Using these algorithms,
    we're able to piece together pictures
  • 6:54 - 6:56
    from this sparse, noisy data.
  • 6:56 - 7:00
    So here I show a sample reconstruction
    done using simulated data,
  • 7:00 - 7:02
    when we pretend to point our telescopes
  • 7:02 - 7:05
    to the black hole
    in the center of our galaxy.
  • 7:05 - 7:09
    Although this is just a simulation,
    reconstruction such as this give us hope
  • 7:09 - 7:13
    that we'll soon be able to reliably take
    the first image of a black hole
  • 7:13 - 7:15
    and from it, determine
    the size of its ring.
  • 7:16 - 7:19
    Although I'd love to go on
    about all the details of this algorithm,
  • 7:19 - 7:22
    luckily for you, I don't have the time.
  • 7:22 - 7:24
    But I'd still like
    to give you a brief idea
  • 7:24 - 7:26
    of how we define
    what our universe looks like,
  • 7:26 - 7:30
    and how we use this to reconstruct
    and verify our results.
  • 7:30 - 7:33
    Since there are an infinite number
    of possible images
  • 7:33 - 7:35
    that perfectly explain
    our telescope measurements,
  • 7:35 - 7:38
    we have to choose
    between them in some way.
  • 7:38 - 7:40
    We do this by ranking the images
  • 7:40 - 7:43
    based upon how likely they are
    to be the black hole image,
  • 7:43 - 7:45
    and then choosing the one
    that's most likely.
  • 7:45 - 7:47
    So what do I mean by this exactly?
  • 7:48 - 7:50
    Let's say we were trying to make a model
  • 7:50 - 7:53
    that told us how likely an image
    were to appear on Facebook.
  • 7:53 - 7:55
    We'd probably want the model to say
  • 7:55 - 7:58
    it's pretty unlikely that someone
    would post this noise image on the left,
  • 7:58 - 8:01
    and pretty likely that someone
    would post a selfie
  • 8:01 - 8:02
    like this one on the right.
  • 8:02 - 8:04
    The image in the middle is blurry,
  • 8:04 - 8:06
    so even though it's more likely
    we'd see it on Facebook
  • 8:07 - 8:08
    compared to the noise image,
  • 8:08 - 8:11
    it's probably less likely we'd see it
    compared to the selfie.
  • 8:11 - 8:13
    But when it comes to images
    from the black hole,
  • 8:13 - 8:17
    we're posed with a real conundrum:
    we've never seen a black hole before.
  • 8:17 - 8:19
    In that case, what is a likely
    black hole image,
  • 8:19 - 8:22
    and what should we assume
    about the structure of black holes?
  • 8:22 - 8:25
    We could try to use images
    from simulations we've done,
  • 8:25 - 8:27
    like the image of the black hole
    from "Interstellar,"
  • 8:27 - 8:30
    but if we did this,
    it could cause some serious problems.
  • 8:30 - 8:34
    What would happen
    if Einstein's theories didn't hold?
  • 8:34 - 8:38
    We'd still want to reconstruct
    an accurate picture of what was going on.
  • 8:38 - 8:41
    If we bake Einstein's equations
    too much into our algorithms,
  • 8:41 - 8:44
    we'll just end up seeing
    what we expect to see.
  • 8:44 - 8:46
    In other words,
    we want to leave the option open
  • 8:46 - 8:49
    for there being a giant elephant
    at the center of our galaxy.
  • 8:49 - 8:50
    (Laughter)
  • 8:50 - 8:53
    Different types of images have
    very distinct features.
  • 8:53 - 8:57
    We can easily tell the difference
    between black hole simulation images
  • 8:57 - 8:59
    and images we take
    every day here on Earth.
  • 8:59 - 9:02
    We need a way to tell our algorithms
    what images look like
  • 9:02 - 9:05
    without imposing one type
    of image's features too much.
  • 9:06 - 9:08
    One way we can try to get around this
  • 9:08 - 9:11
    is by imposing the features
    of different kinds of images
  • 9:11 - 9:15
    and seeing how the type of image we assume
    affects our reconstructions.
  • 9:16 - 9:19
    If all images' types produce
    a very similar-looking image,
  • 9:19 - 9:21
    then we can start to become more confident
  • 9:21 - 9:25
    that the image assumptions we're making
    are not biasing this picture that much.
  • 9:26 - 9:28
    This is a little bit like
    giving the same description
  • 9:29 - 9:32
    to three different sketch artists
    from all around the world.
  • 9:32 - 9:34
    If they all produce
    a very similar-looking face,
  • 9:34 - 9:36
    then we can start to become confident
  • 9:36 - 9:40
    that they're not imposing their own
    cultural biases on the drawings.
  • 9:40 - 9:43
    One way we can try to impose
    different image features
  • 9:43 - 9:46
    is by using pieces of existing images.
  • 9:46 - 9:48
    So we take a large collection of images,
  • 9:48 - 9:51
    and we break them down
    into their little image patches.
  • 9:51 - 9:55
    We then can treat each image patch
    a little bit like pieces of a puzzle.
  • 9:55 - 10:00
    And we use commonly seen puzzle pieces
    to piece together an image
  • 10:00 - 10:02
    that also fits our telescope measurements.
  • 10:03 - 10:07
    Different types of images have
    very distinctive sets of puzzle pieces.
  • 10:07 - 10:10
    So what happens when we take the same data
  • 10:10 - 10:14
    but we use different sets of puzzle pieces
    to reconstruct the image?
  • 10:14 - 10:19
    Let's first start with black hole
    image simulation puzzle pieces.
  • 10:19 - 10:20
    OK, this looks reasonable.
  • 10:20 - 10:23
    This looks like what we expect
    a black hole to look like.
  • 10:23 - 10:24
    But did we just get it
  • 10:24 - 10:27
    because we just fed it little pieces
    of black hole simulation images?
  • 10:27 - 10:29
    Let's try another set of puzzle pieces
  • 10:29 - 10:32
    from astronomical, non-black hole objects.
  • 10:33 - 10:35
    OK, we get a similar-looking image.
  • 10:35 - 10:37
    And then how about pieces
    from everyday images,
  • 10:37 - 10:40
    like the images you take
    with your own personal camera?
  • 10:41 - 10:43
    Great, we see the same image.
  • 10:43 - 10:47
    When we get the same image
    from all different sets of puzzle pieces,
  • 10:47 - 10:49
    then we can start to become more confident
  • 10:49 - 10:51
    that the image assumptions we're making
  • 10:51 - 10:54
    aren't biasing the final
    image we get too much.
  • 10:54 - 10:57
    Another thing we can do is take
    the same set of puzzle pieces,
  • 10:57 - 11:00
    such as the ones derived
    from everyday images,
  • 11:00 - 11:03
    and use them to reconstruct
    many different kinds of source images.
  • 11:03 - 11:05
    So in our simulations,
  • 11:05 - 11:08
    we pretend a black hole looks like
    astronomical non-black hole objects,
  • 11:08 - 11:12
    as well as everyday images like
    the elephant in the center of our galaxy.
  • 11:12 - 11:15
    When the results of our algorithms
    on the bottom look very similar
  • 11:15 - 11:18
    to the simulation's truth image on top,
  • 11:18 - 11:21
    then we can start to become
    more confident in our algorithms.
  • 11:21 - 11:23
    And I really want to emphasize here
  • 11:23 - 11:25
    that all of these pictures were created
  • 11:25 - 11:28
    by piecing together little pieces
    of everyday photographs,
  • 11:28 - 11:30
    like you'd take with your own
    personal camera.
  • 11:30 - 11:33
    So an image of a black hole
    we've never seen before
  • 11:33 - 11:37
    may eventually be created by piecing
    together pictures we see all the time
  • 11:37 - 11:40
    of people, buildings,
    trees, cats and dogs.
  • 11:40 - 11:43
    Imaging ideas like this
    will make it possible for us
  • 11:43 - 11:45
    to take our very first pictures
    of a black hole,
  • 11:45 - 11:48
    and hopefully, verify
    those famous theories
  • 11:48 - 11:50
    on which scientists rely on a daily basis.
  • 11:50 - 11:53
    But of course, getting
    imaging ideas like this working
  • 11:53 - 11:56
    would never have been possible
    without the amazing team of researchers
  • 11:56 - 11:58
    that I have the privilege to work with.
  • 11:58 - 11:59
    It still amazes me
  • 11:59 - 12:03
    that although I began this project
    with no background in astrophysics,
  • 12:03 - 12:05
    what we have achieved
    through this unique collaboration
  • 12:05 - 12:08
    could result in the very first
    images of a black hole.
  • 12:08 - 12:11
    But big projects like
    the Event Horizon Telescope
  • 12:11 - 12:14
    are successful due to all
    the interdisciplinary expertise
  • 12:14 - 12:15
    different people bring to the table.
  • 12:15 - 12:17
    We're a melting pot of astronomers,
  • 12:17 - 12:19
    physicists, mathematicians and engineers.
  • 12:19 - 12:22
    This is what will make it soon possible
  • 12:22 - 12:25
    to achieve something
    once thought impossible.
  • 12:25 - 12:27
    I'd like to encourage all of you to go out
  • 12:27 - 12:29
    and help push the boundaries of science,
  • 12:29 - 12:33
    even if it may at first seem
    as mysterious to you as a black hole.
  • 12:33 - 12:34
    Thank you.
  • 12:34 - 12:37
    (Applause)
Title:
How to take a picture of a black hole
Speaker:
Katie Bouman
Description:

At the heart of the Milky Way, there's a supermassive black hole that feeds off a spinning disk of hot gas, sucking up anything that ventures too close -- even light. We can't see it, but its event horizon casts a shadow, and an image of that shadow could help answer some important questions about the universe. Scientists used to think that making such an image would require a telescope the size of Earth -- until Katie Bouman and a team of astronomers came up with a clever alternative. Learn more about how we can see in the ultimate dark.

more » « less
Video Language:
English
Team:
TED
Project:
TEDTalks
Duration:
12:51

English subtitles

Revisions Compare revisions