The technique that the astronomers of the Event Horizon Telescope (EHT) use to observe black holes is called Very Long Baseline Interferometry, or VLBI, but it might as well be called Extremely Delayed Gratification Astronomy: it can take weeks or months after an observing run to find out whether the telescope array actually saw anything. We’ll get to the reason for the delay in a bit. First, though, some news: the newly upgraded and fielded stations of the EHT are indeed starting to see things.
Now that most of the necessary equipment has been installed in the telescopes that will make up the Event Horizon Telescope array—hand-built, high-frequency cryogenic receivers; brand-new high-bandwidth digital signal processors and recorders that can swallow many terabytes of data per night; atomic clocks; and miscellaneous other crucial bits and pieces that link this stuff together—the trick is to get the telescopes working together. This is a matter of observing simultaneously and looking for common detections between pairs of sites, pair by pair.
When two telescopes in a radio interferometer make a common detection—when both telescopes record the same exact light waves—astronomers say they “found fringes” between the two sites. “Finding fringes” isn’t exactly an intuitive concept, and we’ll get to what it means in just a minute, but first, let me tell you what new fringes the EHT has, so far, found.
1. Fringes between the South Pole Telescope and the Atacama Pathfinder Experiment in Chile. In January, astronomers used the two sites to observe Sagittarius A* and Centaurus A, a radio galaxy 10 million light years from Earth whose nucleus contains a black hole with the mass of 55 million suns. In April, the teams announced that during those January observations the two sites managed to make common detections on both black holes.
2. Fringes to the Large Millimeter Telescope in Mexico.
3. Fringes to the Submillimeter Array (SMA) on Mauna Kea, which a team led by Jonathan Weintroub at the Smithsonian Astrophysical Observatory just equipped with a new correlator—basically, a machine that correlates the signals from the SMA’s small dishes and sums them together as if they came from a single dish.
4. Fringes in January between the Atacama Pathfinder Experiment and Atacama Large Millimeter Array (ALMA), proving that the latter—the world’s largest and most sensitive instrument of its kind—can do Very Long Baseline Interferometry, which it was not initially designed to do.
The term “fringe” is a reference to the pattern of interference formed when two light beams intersect. The concept of the interference fringe goes back to Young’s double-slit experiment. Where those waves are “in phase”—that is, where crests line up with crests and troughs line up with troughs—they add together, producing a stronger signal. Where they are out of phase, they cancel. The result is a series of alternating light and dark ridges.
I called EHT director Shep Doeleman to ask him to explain what, exactly, the term “fringe” means in context of VLBI. I’ll quote his response at length:
“’Fringe’ really is a bit of an historic reference. The way I think about it is generally as a pattern of interference. It’s a coherent merging of light that that tells you something about where that light came from because of the pattern it produces. Shine a laser on two fine slits, and different points on the screen [behind those slits] get light and dark because at some points the path length difference is some fraction of a wavelength. When the path length is the full wavelength, you get constructive interference. When it’s not a full wavelength, you get destructive. What do you learn from that? If you know the distance from the slits to the screen, you can tell how far apart the slits were.
“With VLBI we are trying to find the pattern of interference on the spherical wave front that’s coming to us from the black hole. Think of the black hole as being the double slit; it’s producing interference patterns, or fringes, all the way across the universe to us here on Earth. If we can measure those interference patterns—go around that wavefront and say there’s constructive interference here, destructive interference here—we can put together what that black hole looks like. It’s the shape and the structure of the black hole that made that interference pattern. Basically we collect data from around the world and compare the data and that tells us what that pattern looks like. The whole VLBI network is summing that spherical wave.”
Once you understand what’s involved in summing those waves collected at different sites across the planet, it’s easy to see why it takes so long to get fringes.
“In order to make that measurement, have to know very precisely the geometry of the Earth. You have to line up waves collected, for example, in Chile and California and Hawaii and play them back and make this comparison. When you play them back and see some commonality between signals in Chile and California, then you say ‘Aha, we have fringes.” We were able to verify that the instrument works.”
The reason this is so hard to do: “We don’t know the exact frequency [at which to find the common detections] because the Earth is rotating. One place gets signals slightly Doppler shifted from another place. Plus, there’s delay, because the atmosphere is different in different places. Plus, we don’t know the location of the telescopes on the Earth within a millimeter,” which is the wavelength of the light the Event Horizon Telescope collects. “We know the location of the telescopes within 10 centimeters, because that’s what we get from GPS. Light travels about a foot in one nanosecond. So if we’re off by three or four or five feet, we have to search around before we’ll be able to find out where the fringe is. In the same way, because the Earth is rotating, the frequency shift from site to site can be a few millihertz, or thousandths of a hertz. And delay comes from a lot from different sources. It could be expansion in the structure of the telescope. It could be that we didn’t know where telescope was in the first place. Maybe something was inserted in the beam somewhere.”
Every time I hear VLBI explained in these terms I find myself surprised that the technique works at all. But it does work. And so far, at least, it seems that the newly upgraded telescopes of in the Event Horizon Telescope work, too.