How do you improve a network of radio telescopes that together create a telescope effectively as large as the world? By enabling it to record and process data faster.
Scientists at MIT's Haystack Observatory have moved an important step closer toward that next-generation goal, developing equipment that allows the world-wide network of telescopes (including two at Haystack) to record data faster. A machine to process data faster is being developed and is expected to be in hand by 1996.
Last month Haystack scientists successfully recorded radio signals from a quasar, or disturbed core of a galaxy, at a rate four times faster than is possible with current recording systems for the technique that links the telescopes, which is known as very long baseline interferometry (VLBI).
Increased data rates increase the sensitivity of the telescopes involved, which could lead to better maps of radio sources and ultimately aid our understanding of the physics driving these sources.
VLBI combines data received at telescopes separated by large distances-baselines-to create an image "with finer resolution than each telescope looking alone could ever get," said Joseph E. Salah, director of Haystack.
In VLBI, radio signals are simultaneously received by antennas around the earth and recorded on magnetic tapes. These tapes are then sent to a central facility where they are processed on powerful machines called correlators.
To increase the sensitivity of the technique, scientists need to increase both the rate at which data can be recorded and the rate at which it can be processed. Although there are other ways to increase the sensitivity-such as building larger individual telescopes-increasing the recording and processing rates "is by far the most economical," said Alan R. Whitney, a principal research scientist and assistant director at Haystack.
In the breakthrough last month, the scientists pointed two telescopes (one at Haystack and one near Greenbelt, MD) at a quasar and ran a VLBI test to determine how much data (radio signals) they could record from this source. They achieved a data rate of 1,000 megabits per second, or four times more than any other VLBI recording system in use today.
(The test was not designed to study the quasar. "To do any useful work [on the quasar] you would have to take at least 24 hours worth of data," Dr. Whitney said. "The test was for about three hours.")
The recording breakthrough is a step forward toward the next generation of VLBI technology, but the scientists must also increase the rate at which data can be processed-which turns out to be the bottleneck in the entire VLBI process.
There are currently six correlators in the world-two are at Haystack. Half of these have been designed and built by Haystack scientists.
Each of these correlators is currently capable of processing "more data in one night than the IRS processes in one year," Dr. Whitney said. Nevertheless, he continued, "we can currently record far, far more data than we can process."
So the MIT scientists, in collaboration with a host of other groups around the world, are leading the development of a new correlator, the Mark IV, that will be 100 times as powerful as its predecessor, the Mark IIIA. The computer chip for the Mark IV will be designed at MIT, and MIT will be home for one copy of the new machine, which should be completed by 1996.
ONGOING VLBI RESEARCH
Haystack has been a center of excellence for VLBI for some time, running a variety of experiments that use the technique.
For example Shep Doeleman, a graduate student in physics, and Alan E.E. Rogers, a senior research scientist and assistant director at Haystack, conducted a nine-day VLBI experiment earlier this month to probe the structure of faint quasars. To do the experiment, the scientists took advantage of the newly upgraded Haystack telescope, which allowed them to make observations at a wavelength of three millimeters. (The upgraded telescope is now the largest in the US operating at that wavelength.)
Specifically, the two hope to find evidence for material flow in the disk of matter surrounding the black hole that is believed to exist in quasars. They are currently analyzing the data from the experiment, which came in from nine telescopes around the world, including Haystack, that are linked via VLBI.
Scientists at Haystack also use VLBI to verify and track continental drift. They do so by using the technique to measure the distance between telescopes in the VLBI network. For example, said Dr. Whitney, "we know that the distance between here and the telescope in Germany is increasing by about 1.8 centimeters per year, while Hawaii and the continental US are moving apart by almost eight centimeters per year." He notes that these changes were measured for the first time in the late 1980s.
The technique can currently measure the distance between any two telescopes in the network with an accuracy of about one centimeter. Scientists hope to get that accuracy down to one millimeter.
ALSO AT HAYSTACK
VLBI is only one of four major areas of research at the Haystack Observatory, which is located in Westford, Mass.
The others are radio astronomy, which can help scientists understand the physical conditions in interstellar clouds and the processes of star formation; atmospheric science, which uses radar to explore the structure of the atmosphere at different altitudes; and space surveillance, which uses radar to track satellites and space debris.
The observatory receives its primary financial support from the NSF, NASA, and the Department of the Air Force, as well as other federal agencies and national programs.
A version of this article appeared in the April 28, 1993 issue of MIT Tech Talk (Volume 37, Number 30).