The main disadvantage of radio astronomy compared with optical work is its relatively poor angular resolution. However, in some circumstances, radio astronomers can overcome this problem by using a technique known as interferometry. This technique makes it possible to produce radio images of much higher angular resolution than can be achieved with even the best optical telescopes, on Earth or in space.

In interferometry, two or more radio telescopes are used in tandem to observe the same object at the same wavelength and at the same time. The combined instruments together make up an interferometer. Figure 5.24 shows a large interferometer—many separate radio telescopes working together as a team. By means of electronic cables or radio links, the signals received by each antenna in the array making up the interferometer are sent to a central computer that combines and stores the data. The technique works by analyzing how the waves interfere with each other when added together. (Sec. 3.1) If the detected radio waves are in step, they combine constructively to produce a strong signal. If the signals are not in step, they destructively interfere and cancel each other. As the antennas track their target, a pattern of peaks and troughs emerges. After extensive computer processing, this pattern translates into a high-resolution image of the target object.

Figure 5.24 This large interferometer is made up of 27 separate dishes spread along a Y-shaped pattern about 30 km across on the Plain of San Augustin in New Mexico. The most sensitive radio device in the world, it is called the Very Large Array or VLA, for short. (b) A close-up view from ground level of some of the VLA antennas. Notice that the dishes are mounted on railroad tracks so that they can be repositioned easily.

An interferometer is essentially a substitute for a single huge antenna. As far as resolving power is concerned, the effective diameter of an interferometer is the distance between its outermost dishes. In other words, two small dishes can act as opposite ends of an imaginary but huge single radio telescope, dramatically improving the angular resolution. For example, resolution of a few arc seconds can be achieved at typical radio wavelengths (such as 10 cm), either by using a single radio telescope 5 km in diameter (which is impossible to build) or by using two or more much smaller dishes separated by 5 km and connected electronically. The larger the distance separating the telescopes—the longer the baseline of the interferometer—the better the resolution attainable. Large interferometers like the instrument shown in Figure 5.24 now routinely attain radio resolution comparable to that of optical images. Figure 5.25 compares an interferometric radio map of a nearby galaxy with a photograph of that same galaxy made using a large optical telescope. The radio clarity is superb—much better than the radio contour map of Figure 5.23.

Figure 5.25 (a) VLA radio "photograph" (or radiograph) of the spiral galaxy M51, observed at radio frequencies with an angular resolution of a few arc seconds (a); shows nearly as much detail as (b) an actual (light) photograph of that same galaxy made with the 4-m Kitt Peak optical telescope.

Astronomers have created radio interferometers spanning very great distances, first across North America and later between continents. A typical very-long-baseline interferometry experiment (usually known by the acronym VLBI) might use radio telescopes in North America, Europe, Australia, and Russia to achieve angular resolution on the order of 0.001", about 1000 times better than images produced by most current optical telescopes. It seems that even Earth's diameter is no limit. Radio astronomers have successfully used an antenna in orbit, together with several antennas on the ground, to construct an even longer baseline and achieve still better resolution. Proposals exist to place interferometers entirely in Earth orbit, and even on the Moon.

Although the technique was originally developed by radio astronomers, interferometry is no longer restricted to the radio domain. Radio interferometry became feasible when electronic equipment and computers achieved speeds great enough to combine and analyze radio signals from separate radio detectors without loss of data. As the technology has improved, it has become possible to apply the same methods to higher-frequency radiation. Millimeter-wavelength interferometry has already become an established and important observational technique, and infrared interferometry will become commonplace in the next few years.

Optical interferometry is the subject of intensive research. In 1997 a group of scientists in Cambridge, England, succeeded in combining the light from three small optical telescopes to produce a single, remarkably clear, image. Each telescope of the Cambridge Optical Aperture Synthesis Telescope (COAST) is only 0.4 m in diameter, but when the equipment is positioned 6 m apart, the resulting resolution is a stunning 0.01". This enables a "splitting" of the binary star Capella, whose two member stars are separated by only 0.05" and which therefore is normally seen from the ground only as a slightly oblong blur. With the new Cambridge telescope array, the two stars are cleanly and individually separated. Even HST, operating above the atmosphere in orbit, cannot match these test observations with this brand-new device. The Keck telescopes on Mauna Kea are also designed to be used for infrared—and perhaps someday for optical—interferometric work.