Radio Sociology
By George W. Swenson, Jr.
Or should we have called it "Upside-down Radio Astronomy"?
On October 2, 1970, William W. Cochran of the Illinois State Natural History Survey called from Southern Illinois and asked if I could get an airplane and fly down there early the next morning to try to locate a migrating, radio-tagged, sharp-shinned hawk he'd been tracking. Bill is a pioneer in the art of radio telemetry of wildlife, having designed many of the techniques in common use for locating and monitoring wild animals and birds. Once in a while one of his subjects would move out of radio range of his direction-finding truck and disappear; then it was sometimes possible to re-acquire the signal from a light airplane. So, the next morning I showed up at Carmi, Illinois, in a Piper Arrow, after a predawn flight from Urbana.
The 222 MHz receiving antenna was a four-element Yagi normally mounted on a rotating mast atop the truck. We took it down and clamped it to the entrance step on the right side of the airplane so that it pointed in the direction normal to the direction of flight . The idea was to climb to an appropriate altitude and fly in large circles to scan a substantial area for the missing bird. Early in the morning the hawk would be roosting somewhere, waiting for the atmospheric thermals to develop before continuing its southward flight.
We took off and commenced the routine, making gradual turns while listening intently with headphones. While turning to the right, the plane banks to the right, so the antenna points somewhat below the horizon, its beam sweeping across the ground at a shallow angle. For a while there was nothing to be heard; then I noticed that when the antenna beam swept across the small town of Olney (pop. 8600) there was a rush of noise-like sound in the headphones. Bill confirmed this, so we looked at a couple of other towns, all of which had similar noise signatures. It was apparent that concentrations of population correspond with areas of increased noise temperature* . This was no real surprise, as we radio astronomers had long known that we couldn't make sensitive observations from within or near cities or large towns, but I was intrigued by the very obvious effect we observed.
* The power received in bandwidth B is given by P = KTB, where K is Boltzmann's constant and T is the equivalent noise temperature.
Copyright (c) 1996 Institute of Electrical and Electronics Engineers. Reprinted from the IEEE Antennas and Propagation Magazine, vol. 38, pp.45-47, April 1996.
This material is posted here with permission of IEEE. Internal of personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution must be obtained from the IEEE by sending a blank e-mail message to info.pub.permission@ieee.org.
From chasing the hawk to chasing funds
We found the hawk, roosting in a tree right on the airport at Henderson, Kentucky. I flew back to Urbana, leaving Bill to follow the bird in his truck.
On the way home I speculated on the sources of the radio noise. Over the years we'd experienced much trouble with man-made interference at our Vermilion River Observatory, usually being able to identify the source: local electric fences, automobile ignition, weather balloon-sondes, out-of-band spurious emissions from TV transmitters, etc. Other radio astronomers continually experience similar problems. But here was a phenomenon different in that the noise was broad-band and featureless, like the "signal" from a strong cosmic source as sensed by a radio telescope. The radiation might be indistinguishable from that of a "black body" and thus could be characterized by its "equivalent temperature." It would be interesting to study the relationship between a town's radiation signature and its other attributes such as economic type, population, geographic concentration, etc.
I sought out an urban sociologist at the Chicago campus of the University of Illinois, who greeted the proposition with enthusiasm. Together we put together a proposal to the National Science Foundation, who responded quite promptly that they did not, at that time, support social research. Undeterred, we sent the proposal to the Ford Foundation, who replied that this was physical science, which was outside their domain. Later I learned that the NSF proposal had resulted in a "cross-cut review" among government agencies, who had agreed that none of them would support this work. This puzzled me at first, as I hadn't viewed the matter as being of that great import. However, it subsequently appeared that certain agencies were researching the phenomenon themselves, in classified programs. By now somewhat frustrated, I abandoned the sociological aspect and submitted a much scaled-down proposal to the University's Research Board, who granted a modest sum for airplane rental and some electronic components.
Bill Cochran and I rigged one of the University's training planes, a Cessna 150, with some downward-looking Yagi's for 148, 222, and 440 MHz and a trailing dipole for 74 MHz. We built low-noise converters for those frequencies and modified a communication receiver to suppress its output during the airplane's ignition noise impulses. Data were recorded on a Rustrak strip-chart recorder, along with noise calibration signals from a semiconductor source. On the first test flight we confirmed that an open cornfield gave an equivalent noise temperature of 300 K, a comforting confirmation that the system worked correctly. On that flight it was necessary to demonstrate that the ignition-noise blanker did not bias the measurement. This was done by deliberately turning the ignition switch off during flight and stopping the engine completely, an uncomfortable and never-repeated procedure that presented a challenge to years of aeronautical training and practice. In each of the test frequency bands a specific frequency free from transmitter signals was selected so that only broad-band, unintentional noise would be recorded. Only one Yagi at a time could be carried, so there was much landing and taking-off to change antennas.
Over the cities and the small towns
The first measurements were to try to find the principal sources of noise. We flew over a large coal-fired power generating station, high voltage transmission lines, interstate highways, and TV stations. None gave significant noise. Only towns with significant concentrations of infrastructure were observable. We then adopted a standard route, from west of Champaign, along US Highway 150 eastward to Danville and on to the Indiana state line, and return, all at 700 meters above the ground. Flying this route many times at all hours of the day and all days of the week, I became so familiar with the noise signatures of the several towns that I could navigate under a hood with no other position reference, merely by watching the strip chart. This suggests, possibly, why the government was not enthusiastic about our program. The smaller towns were not resolved by the antenna beams so they were treated as point sources. The two larger cities, Champaign-Urbana and Danville, were resolved, so their contributions could be characterized by surface temperatures. Over many flights we saw consistent results. The temperatures of the larger cities showed strong diurnal variations, presumably related to human activity. Above 74 MHz the surface temperatures were inversely related to frequency; non-thermal, as astronomers would say.
The city surface temperatures varied typically from 9600 Kelvin at 148 MHz to 2400 K at 440 MHz. Our purpose at this point was to provide a benchmark against which possible deterioration with time of the electromagnetic noise environment could be measured. The results of these first measurements were published in Science (vol. 181, 543-545, 1973).
The smaller towns and hamlets were harder to characterize, though they generally conformed to the non-thermal model. The little village of Fithian (pop. 532) always showed an anomalously high noise spike, especially at 222 MHz; this demanded a closer investigation. We drove there with a battery-powered receiver and walked around the streets, pointing the Yagi at houses and other possible sources and acquiring a retinue of young boys and dogs as we went. Finally it became apparent that the source was in the garage of a house on a spacious lot. With some trepidation we approached the house. Fortunately, the occupant, who readily admitted us to the garage, was an employee of the University's power plant, somewhat accustomed to professors and their foibles. The noise source turned out to be the superregenerative receiver of an automatic garage-door opener, an old vacuum-tube model whose squegging detector radiated nonchalantly over a very broad band. I think many of the spikes we saw pushing up through the noise plateaus of the cities were caused by these primitive receivers.
Fast forward to 1988
And so it stood for sixteen years. Then, in 1988, I thought it time to see if there had been any secular changes in the urban noise environment, and graduate student Mark Cudak needed a thesis topic. Bill Cochran was still present with his practical expertise. This time the NSF was sympathetic, enthusiastic, even. The country's large investments in radio astronomical facilities had heightened concern about radio spectrum pollution. Now it was possible to employ some more sophisticated technology: a laptop computer for data recording and control, commercial noise calibration sources, and narrowband r-f filters to guard against intermodulation from nearby TV stations. My flying club had a Cessna 150 just like the previous one. This time we built duplicate receivers so we could carry two antennas at once, a downward-pointing log-periodic dipole array for 412 and 222 MHz under the right wing and a folded dipole on 148 MHz under the left wing, the wing serving as a reflector. The antenna patterns were measured by flying over target transmitters. Fellow pilots of the Enginaires Aero Club entered into the spirit of things with enthusiasm, so the burden didn't fall entirely on me this time.
Mark quickly discovered that the laptop computer made much more noise than the target sources, so he had to enclose it completely in a metal box with heavily filtered leads, leaving accessible only a single push-button to cycle the programs. Imprecise navigation had proved to be a major source of measurement error, so we tried a Loran-C receiver to improve matters. This, too, proved to be overwhelmingly noisy, so that effort was abandoned.
Optimistic results
We repeated the earlier campaign of measurements, this time with many more flights and more complete seasonal and diurnal coverage. Mark analyzed the data and we published it in Radio Science (vol. 26, 773-781, 1991). I had expected to find the cities more noisy than they were 17 years earlier; computers, automobiles, and all kinds of electric appliances had proliferated in the interim. To everyone's surprise, I think, the situation had not changed much, if any, over the years. The spectrum of the cities had the same slope, varying from 9000 K at 148 MHz to 1600 K at 412 MHz. The midday noise temperatures at Champaign were two dB lower than in 1972, about the same within experimental error. A composite of the two spectra is shown in Figure 1. The lack of increase in noise pollution, despite the increase in electrical apparatus, is a hopeful sign that regulation of spurious emissions from apparatus is having the desired effect. Of course, this experiment was confined to a restricted venue; still Champaign-Urbana is a community with a very high concentration of computers and "high-tech" activity of various kinds. (In 1989 we did not see so many discrete noise spikes as before, suggesting that newer models of garage-door openers, in particular, have cleaner receivers.)
This somewhat optimistic result should not reduce our concern for spectrum pollution and conservation. The proliferation of transmitting services continues apace, and we should be cautious in our exploitation of this precious resource.
Figure 1. The secular peak midday surface brightness temperature variations of Champaign-Urbana.
Figure 2. West-to-east flight record of June 28, 1989, 12:26 pm, 222 MHz.