Astronomers Capture First Image of a Black Hole, National Science Foundation
“The Event Horizon Telescope (EHT) — a planet-scale array of eight ground-based radio telescopes forged through international collaboration — was designed to capture images of a black hole.”

Marc Boucher is an entrepreneur, writer, editor, podcaster and publisher who has worked in space media for more than 25 years. In 1999 he co-founded SpaceRef with Keith Cowing. The new media company grew into a network that included NASA Watch and The Astrobiology Web. After selling SpaceRef, he joined Multiverse Media, which later merged with SpaceNews. There he helped steer the combined company toward more multimedia, a subscription paywall and intelligence products, and pushed for a mobile app. He is also the founder of SpaceQ Media, Canada’s space news and market intelligence publication.

Marc started his first Internet content business in 1992 and hasn’t looked back. When he’s not working, he loves exploring Canada and the world and expanding his photography skills.

31 replies on “Researchers Reveal First Image of a Black Hole”

  1. Ultra cool.
    I am surprised that the image was of the one in M87 instead of Sgr A* which is at the center of our galaxy. I would a expect a Sgr A* image would show more detail.

    1. I imagine that the M87 image was captured partially due to our stellar viewpoint being from above (or below) that galaxy’s disk; so it’s unobstructed. Unfortunately, for our home galaxy, we’re stuck with an edge-on view with several billion stars annoyingly blocking us.

      OTOH, given the obscene amount of radiation and cosmic rays those monsters produce – especially from the poles – I’d consider that “annoyance” to be more of a nice insulating wall.

      1. Yes, that is why some researchers believe that galaxies have habitable zones just like stars, and if a star is too close to the center it won’t have planets with life on the surface.

        https://www.astrobio.net/ne…

        Galactic Habitable Zones

        1. Reminding me of some work by Vernor Vinge, in which he imagines the speed of light variable by ‘zones’ within the galaxy. He also imagines intelligent dogs, as well as spiders, the latter of which I didn’t realize until too far into the story for my arachnophobia to kick in.

          1. It has nothing to do with the speed of light. It merely recognizes that the radiation a planet is exposed to varies greatly based on its location within a galaxy. Really just common sense stuff if you are trying to determine where life may be able to exist, no different than considering temperature and presence of an atmosphere. But it seems to be ignored by the SETI folks.

    2. M87’s black hole’s Schwarzschild radius is about 78% of the angular size of Sagistarius A*’s Schwarzschild radius. Plus, the view into Sagistarius A* is through its accretion disc. Given that, it’s probably “easier” to see M87’s black hole thus making it a better first target.

      1. Viewing through the accretion disk doesn’t matter. The gravitational field is so strong that the light is still bent so much that you still see what’s on the back side. You would just see something different than a ring. Think of the view from the movie “Interstellar”.

        Here is a good explanation of how the viewing geometry effects what you see.
        https://www.youtube.com/wat…

        After posting my initial post I read that they are still processing the data for Sgr A* but it’s more challenging because it’s rotating a 1000x faster.

    3. The BBC report mentioned that M87’s back hole is larger and brighter than our galaxy’s. They didn’t say by how much, and also mentioned the reason for this was not understood. One of the researchers was quoted as saying, M87’s black hole was “one of the heaviest black holes that we think exists. It is an absolute monster, the heavyweight champion of black holes in the Universe.”

        1. I’m just quoting the BBC, who were quoting a Professor Heino Falcke. I don’t know about him, but it isn’t unusual for scientists to get the “don’t use jargon; don’t say anything a sixth grade student wouldn’t immediately understand,” lecture from media relations people.

  2. So, VLBI is pretty cool. Can we do something similar for other wavelengths / other targets?

    This article says EHT gets a resolution of 20 micro-arcseconds. WA tells me Proxima Centauri is 1.9 milli-arcseconds, about 100 times bigger than EHT’s resolution. Would this same equipment be able to discern surface features on Proxima, or is Proxima not interesting at the detected wavelength?

    If you could apply this to other wavelengths, you might even be able to directly image exoplanets, although atmospheric distortion might mean that you would want a space- or moon- based group of telescopes (Fleet of telescopes? Herd?) for other wavelengths.

    1. Currently, we can’t manage shorter wavelengths. You need knowledge of the telescopes’ relative position to better than a wavelength and mix the signals together with timing accuracy of better than a wave period. Optical interferometry that with mirrors, mixing the light directly. But the telescopes have to be relatively close together for that. For these observations (submillimeter radio), the signals were coherently amplified and they recorded the actual wave forms and mixed them digitally. (Recoding that, by the way, took a huge amount of data volume. This work, and others like it, don’t even try sending the data over the internet. They record it on a truck load of hard drives and ship the hard drives to a central location.)

      So that’s the challenge for doing this sort of thing at shorter wavelengths. Knowing the position of the telescopes sufficiently well, recording the signals at sufficient time resolution and accuracy, and dealing with all the data that implies. I think this work is pushing the state of the art.

      Correction: The knowledge of the telecsopes’ positions doesn’t really have to be good to better than a wavelength. There are some geometric factors and they can be large. I should have written that the positions have to be known precisely and the required level of precision scales with the wavelength.

      1. How are the radio wavelengths translated into those of color? Is there a simple 1:1 substitution? For instance, SHF1—>Color1; SHF2—>Color2, etc. ? Perhaps this sort of color mapping is similar to that sometimes used on the Cassini images for both scientific and aesthetic purposes?

        1. In this case, I think, as others have noted, that it’s just a grey scale overlaid with a orange-to-yellow hue. That’s not uncommon in the field. Adding the changes in hue makes it easier to spot the more subtle features in the image. Sometimes the color (it could be green-blue, purple, or just about anything) are simply an esthetic choice by the scientists involved. Sometimes it does mean something if you know the code. Solar physicists have half a dozen emission lines they regularly monitor, and they’ve actually managed to agree on a color convention. If you see one of those black and white images at a ultraviolet wavelength displayed with a green rather than a yellow tint, that’s a code for which one of those emission lines it is.

          But this can also be a problem, and yes, scientists do argue about color tables. For spectrograms or contour plots of theoretical calculations, there is a very popular one that overlays the grey scale with a rainbow going from black (at 0 on a scale of 0 to 255), to blue, to cyan, to green, to yellow, to orange to red. People like it for esthetic reasons, because it makes the figures look more dramatic. But it also produces optical illusions. The cyan in particular will look like very sharper contrast to neighboring values, so people do see (or think they see) a sharp boundary around 85 on a scale of 0 to 255. Worse, that color table makes figures more or less incomprehensible to people who are color blind. There are other color tables specifically designed not to produce optical illusions, as well as being more comprehensible to someone who is color blind. (I’ve been told it’s no worse than if it were an uncolored greyscale image.) Those haven’t caught on because they tend to be more muted and not as visually exciting. And, along these lines, people have fiddled with color tables to “pull out” the feature they want the audience to notice. That’s frowned upon, since it’s too easy to pull out something that isn’t really there.

          For real, multi-wavelength images, usually people just assign one wavelength or filter to blue, a second to green and a third to red. Sometimes the choice is purely esthetic. There are six permutations, and you can just try them all and use the one that looks best. Sometimes, people try to put the wavelengths in the right order (i.e. they assign the shortest wavelength to blue and the longest to red.) I’ve even seen someone produce what he called a “true color infrared image.” He used infrared wavelengths which had the same wavelength ratios as the red, green and blue in the image. In other words, it was “true color” if you had been observing at a huge doppler blue shift which put those wavelengths into the visible range. That was an esthetic failure, and basically only proved that he was a serious nerd with way too much free time…

    2. The problem is that EHT is only operating in the roughly 200-300ish GHz range (At two specific frequencies which I don’t remember. I’ll bet exoplanet emissions in that spectrum are zero or virtually zero.

      1. Probably. That’s unfortunate, since that’s an interesting part of the spectrum. ALMA has done some very impressive planetary science at those frequencies. But the flux from planets in our solar system is much, much higher than it would be from an exoplanet. On the other hand, there are some interesting things you could do in terms of stellar physics. The Sun emits at those wavelengths during flares and other energetic-particle-producing events. Imaging that on another star would definitely be interesting. Unfortunately, EHT was about a $50 million dollar project, and I doubt you could get that kind of money to look at flares on Proxima Centauri.

      1. Yes and no. The VLT is an interferometer, but it’s an optical one. The telescopes are only, what, 50 meters or so apart? You can mix the light before it reaches the detector. Radio interferometry with telescopes all over the world requires recording phase information at every site and then mixing the signals in post-processing. I think that’s possible up to a few hundred GHz or possibly a THz (1-0.3 millimeters). But doing that all the way down to near IR or visible wavelengths is beyond the state of the art.

        1. I beg to differ. The VLT does operate in the visible and infrared and does what is being done with radio but on a much smaller scale. An interferometer is an interferometer regardless of the wave length it’s working at.

          1. Those are really, fundamentally different measurements. The VLT and other interferometers at those wavelengths works by combining the light _before_ measuring it. Whether it’s by mirrors, waveguides, or whatever, the signal is sent to a central location and combined to produce the interference pattern before any measurement is made. Then they measure the interference pattern. In the radio and sub-millimeter, they record the signals at every telescope, separately, and with extremely high time resolution. Then they take the measurements, not the light itself, to a central location and use a computer to combine the data from the various telescopes. Those differences place fundamentally different limits on what is and is not possible.

            For the optical systems, you can’t have your telescopes thousands of kilometers apart. Actually, you can’t have them one kilometer apart. If you tried, you couldn’t get the signals to a central location with mirrors or whatever. That limits their resolution.

            For the radio frequency (heterodyne) you need, at the very least, a good, high quality local electric oscillator at the frequency you are observing. Those don’t exist above 1 THz or so (about a third of a millimeter wavelength). That limits the frequency/wavelength.

            Current technology does not allow the wavelengths observable by optical systems _and_ the thousands of kilometer baselines of the heterodyne systems.

  3. One of my gripes is that astronomical images like this go out to the public without any mention that the colors are artificially assigned. I’m already seeing comments on various sites suggesting – incorrectly – that the ring appears red-orange due to gravitational red shift. The correct answer is that the researchers chose to represent intensity (not frequency) of the collected radio signals using a common color scale where black-red-orange-yellow-white denotes increasing intensity.

    1. It’s annoying, and I suspect you’re right about the colors. I actually recognize the color table. But I worry more about the Juno images of Jupiter. The project, or people on the science team, frequently shows them without mentioning how heavily processed they are. And the processing is entirely for esthetic value. That’s caused questions about things like comparisons with the images from the Cassini flyby. The poles are blue relative to the equatorial clouds, and the contrast is much more striking in the Juno images. That’s caused some people to ask what changed between Cassini’s 2001 encounter and Juno getting there in 2016. Nothing. The Cassini images processed to be as close to true color as possible. The Juno images were processed to make them look cool and dramatic. I’m also worried about students who get excited about astronomy when they see those heavily processed images, and then disappointed and discouraged when they start taking courses in astronomy and see what Jupiter really looks like.

      1. “I worry more about the Juno images of Jupiter.”

        In the mode of being thankful for the small things in life, as my mother insisted, there is satisfaction to be found when the un-scienced view the ‘pictures.’

        1. Please don’t misunderstand. I really do like those images. But they are art, not what Jupiter really looks like. That’s fine. I know an artist who does similar work, although the raw images come from a microscope more often than a telescope. But they are shown (and occasionally sold) as art, not presented at scientific conferences or with any claim they are real, life-like depictions of something.

          I just wish they would occasionally show a true-color, unprocessed image as well, or mention that the images are so heavily processed more often. I’m also worried that someone will try to use them as scientific measurements. If you tried to work out how sharp a hurricane’s eye wall was from the images, the heavy use of edge-enhancement algorithms in Photoshop would be a problem. I’ve got nothing against art, and I especially like the Juno image processed to be a background, with van Gogh’s Starry Night as the foreground. I just think people should be clear about what is art and what is true to life.

Comments are closed.