Tunnel in the sky
To unpack: the “Array” Wootten mentions eventually grew into ALMA, but the really significant word in that sentence is “Millimeter.” Millimeter (one thousandth of a meter, or 1/25 of an inch) is a measure of wavelength, and wavelength is one of the things that astronomy is all about.
Like a snow-bound hermit on the first spring day, astronomers over the last century have, one after another, thrown open a series of new windows on the universe. For most of human history, we thought that everything out there consisted of only the stuff we could see—the stuff that shone in, or reflected, visible light. Slowly we learned that there is much more to the picture than that. Radiation all up and down the electromagnetic spectrum, of which visible light is only a portion, rains down on Earth in the form of “photons,” discrete packets of energy, every second of every day.
Photons come in different colors, both visible and (if you will) invisible. It’s the wavelength of the light that makes those differences. Wavelength is the difference between blue and green and yellow and orange and red, and it’s the difference between each of the invisible “colors” that make up the rest of the electromagnetic spectrum—from the ultra-high energy gamma rays, with wavelengths the size of the nucleus of an atom, out to radio waves, which can have wavelengths the size of a football field.
But there’s a problem. A lot of this energy never makes it through our atmosphere. Air is largely transparent to visible light, and to the longer species of radio waves, but it’s very efficient at blocking most of the rest, including the millimeter and submillimeter waves that ALMA looks at. If the atmosphere were as opaque to visible light as it is to millimeter light, the sky on the clearest of days would only give off a dim gray glow heavily tinged with red and we would have no idea that there were any stars at all.
Living beneath a giant blind spot, astronomers had only a limited sense of how much was going on in the millimeter universe. A few instruments and experiments aside, it wasn’t until the launch of a satellite called IRAS, designed to look at broad range of infrared radiation, that astronomers had any real idea what they were missing. What IRAS found by sticking its nose out from under the enveloping comforter of our atmosphere was truly staggering: hundreds of thousands of new astronomical objects, including 75,000 starburst galaxies blazing away at wavelengths that had not fully observed before. According to Wootten, “for the first time we were able to see the far infrared sky and discover that there were more photons there than there were in any other wavelength range.” It was a little like opening the door of your dismal, sepia Kansas farmhouse and discovering the Land of Oz outside in glorious technicolor.
If you want to build a big millimeter array, only a place that’s high and dry will suit, because it’s the moisture in the atmosphere that defeats millimeter waves. The Chajnantor Plateau, the spot Bob Brown discovered in Chile, three miles up in one of the driest deserts on earth, is as good as it gets. Al Wootten recalled the testing: “In 1995, we put some solar-powered equipment on the site to measure conditions, and we thought ‘Wow, this is spectacular.’”
There is an invisible universe out there waiting to be observed—violently exploding, or gently seething, with energies that carry clues to things we cannot yet imagine. ALMA, at 16,500′, is a tunnel through the sky, a way for us to place our ears right up to the verge of the universe to overhear its secret murmurs.
Business time
The star HD 142527 is so faint and nondescript that it doesn’t have a name, merely a label derived from its place in a catalogue. But it was a desirable target as a part of ALMA’s first period of observing because it is known to be a young star, with a large disc of dust and gas still in the process of forming into a mature solar system. A faint hiss of millimeter radiation that left the star system 450 years ago, around the time of the birth of Galileo, hit the telescope during its trial research run in 2012.
The surface of each antenna dish is engineered out of aluminum panels to form a gentle parabolic curve precise to 20 millionths of a meter, or about 1/5 the thickness of a sheet of paper. The whisper of radiation gathered by the 1,200-square-foot collecting area of each dish is reflected upward and focused toward a secondary reflecting surface, which in turn bounces and focuses the energy down through a hole in the center of the dish, where the telescope starts to get down to business.
The structure of the dish serves only as a very large bucket to gather photons and funnel them to the radio receiver. There are a number of receivers in each telescope, each tuned to a specific wavelength range that is of interest to astronomers. The radiation is so faint, and the receivers so exquisitely sensitive, that even the negligible radiation given off by the equipment itself at normal temperatures is enough to overwhelm the signal. So the receiver cartridge, the “front end” as astronomers call it, has to be cooled to -450 degrees Fahrenheit. Only with the atoms that comprise it quieted down to within a few degrees of absolute zero, is the noise generated by the equipment itself low enough that ALMA can hear the signals it’s designed to gather.
Half of ALMA’s receiver cartridges were assembled at the North American ALMA Integration Center—part of the NRAO Technology Center here in Charlottesville, which occupies a few nondescript buildings that used to be part of the Institute of Textile Technology behind Boxwood Estate on Ivy Road. The buildings may not be monumental, but the work that is done there is world-class. Mark Adams, spokesman for the NRAO director’s office, called the Technology Center staff “an incredible cadre of talent.” They develop and build, painstakingly and by hand, many of the cutting-edge radio frequency amplifiers and superconducting components used in radio astronomy instruments around the world.
From the telescope front end, the amplified signal is now strong enough that it no longer needs supercooling. It is shunted to the back end, where is digitized, and begins the process of being synchronized with the data from each of the other antennas. Buried fiber optic cables carry it along a path whose length is monitored and maintained to a precision of 1 micron—one millionth of a meter. Without that precision, the next stage, the assembly of the data from all the telescopes, would be impossible.
The ALMA Correlator, the brains of the whole operation, was also born in the same humble surroundings. Joe Greenberg, senior engineer, was part of the team that built the Correlator. He designed the custom microchips and the circuit boards that hold them, and this is not his first rodeo. He has been working on correlators since he started at NRAO 40 years ago (minus a stint in the private sector).
The Correlator turns a collection of individual telescopes into something called an “interferometer”—in effect, it orchestrates the information from each antenna and uses it to build a virtual telescope that is as sensitive as all of the antennas combined and as large as the distance between them. So if the array is spread over the full 11 miles of its capacity, it becomes, in effect, an 11-mile-wide telescope. It does this by some sophisticated algorithms, and by just plain brute force number crunching. Greenberg walks through the math: “There are 4,096 processors on each chip, 30 chips per board, and 512 boards in the Correlator. The whole thing runs at 125 Megahertz, which means that each part cycles 125 million times per second.” That means the ALMA Correlator performs 17 quadrillion operations per second. NRAO calculates it would have taken $1 billion worth of commercial computer equipment to do that work. Greenberg and his Technology Center colleagues built it for $11 million.
In the near future, a typical day at ALMA, with the whole array up and running, will see it perform many hours of observing and produce up to 800 gigabytes of data. That’s 200 single-sided DVDs every day, 6,000 per month, 72,000 per year, 2.2 million over the projected 30-year lifespan of ALMA. You can see why Al Wootten said his main job these days is “playing traffic cop.” “Right now,” he said, “the push is, how do we process this data? How do we turn it into images? How do we get it to the scientists? How do we help them produce the science with the least impediment?”