SP-4218 To See the Unseen


- Technical Essay: Planetary Radar Astronomy -



[275] The basic technology of planetary radar astronomy is, as the name implies, radar. Radar is an acronym for RAdio Detection And Ranging. U.S. Naval officers Lieutenant Commanders F. R. Furth and S. M. Tucker devised the acronym in 1940. By 1943, all allied forces had adopted the name, though it remained a classified term until after the second world war, when the acronym radar received general international acceptance, though more as a term than as an acronym.1

As the expression "radio detection and ranging" denotes, radar involves the use of radio for both detection (is it there?) and ranging (how far away is it?). Radar involves transmitting electromagnetic waves (commonly known as radio waves) toward a target and receiving the echoes from that target.

The wavelength of a radio or radar signal traveling through space is measured in meters and fractions of a meter (decimeter, centimeter, millimeter), and its frequency, that is, the number of waves per second, is expressed in hertz or multiples of hertz. One hertz is one wave per second. The high-frequency radar waves used in planetary research are expressed in megahertz (MHz, a million hertz) and gigahertz (GHz, a billion hertz).

There is no real difference between radio and radar waves. International treaties and regulatory agencies have set aside certain groups of radio frequencies, called bands, for specific radio uses, including radar applications (see Tables 11 and 12). Although the first radars to attempt detections of the Moon and Venus operated in the UHF band, planetary radar astronomy today uses only the S and X bands.

A radar system consists of a transmitter and a receiver, plus modulators, signal processors, and data processors. Generally, the transmitter and receiver share the same antenna. Such an arrangement is called a monostatic radar. When the transmitter and receiver do not share the same antenna, that is, when they are located in different places, it is called a bistatic radar.

All of the radars used since 1958 to study solar system objects have a parabolic or dish shape, with one exception. The exception is the Arecibo antenna, which is spherical.

Another difference among the planetary radars is their transmitters. The Millstone Hill radar, which Lincoln Laboratory investigators used to attempt a detection of Venus as early as 1958, was a pulse radar. Pulse radars transmit short bursts of energy and are best suited to tracking objects. Millstone, in fact, was an experimental prototype of a Ballistic Missile Early Warning System (BMEWS) radar used to detect and track potential incoming enemy missiles. In contrast are the continuous-wave radars. They transmit a continuous flow of energy and are better suited for communications applications. JPL's planetary radars at Goldstone are continuous-wave radars.

The output power of radars is expressed in watts. When comparing the power of pulse and continuous-wave radars, one must keep in mind that although pulse radars have relatively high peak power outputs (that is, the amount of power at the highest part of the pulse), their average power output, a measure more comparable to that of the continuous-wave radars, is much lower. Average power is what counts. Thus, while the Millstone pulse radar had a peak transmitting power of 265 kilowatts in 1958, the JPL transmitter in...


[276] Table 11. Radar Frequency Bands and Usage.







3-30 MHz

Over the horizon radar


30-300 MHz

Very-long-range surveillance


300-1000 MHz

Very-long-range surveillance


1-2 GHz

Long-range surveillance

Enroute traffic control


2-4 GHz

Moderate-range surveillance

Terminal air traffic control

Long-range weather


4-8 GHz

Long-range tracking

Airborne weather detection


8-12 GHz

Short-range tracking

Missile guidance

Mapping marine radar

Airborne weather radar

Airborne intercept


12-18 GHz

High-resolution mapping

Satellite altimetry


18-27 GHz

Little used (water vapor)


27-40 GHz

Very-high-resolution mapping

Short-range tracking

Airport surveillance


40-110 GHz

Smart munitions

Remote sensing


110 GHz+


Remote sensing

Source: Fred E. Nathanson, Radar Design Principles, 2d ed. (New York: McGraw-Hill, 1991), p. 19.


[277] Table 12. Standard Radar Frequency Bands.







3-30 MHz

None assigned (in practice, from just above the broadcast band, 1.605 MHz, to 40 MHz or higher)


30-300 MHz

138-144 MHz

216-225 MHz


300-1000 MHz

420-450 MHz

890-942 MHz (at times included in the L band)


1000-2000 MHz

1215-1400 MHz


2000-4000 MHz

2300-2500 MHz

2700-3700 MHz


4000-8000 MHz

5250-5925 MHz


8-12 GHz

8.5-10.68 GHz


12-18 GHz

13.4-14.0 GHz

15.7-17.7 GHz


18-27 GHz

24.05-24.25 GHz


27-40 GHz

33.4-36.0 GHz


40-75 GHz

59-64 GHz


75-110 GHz

76-81 GHz

92-100 GHz


110-300 GHz

126-142 GHz

144-149 GHz

231-235 GHz

238-248 GHz

Source: Fred E. Nathanson, Radar Design Principles, 2d ed. (New York: McGraw-Hill, 1991), p. 18.


...1961 was more powerful, though its average power output was only nine kilowatts. When the current Arecibo upgrade is completed, it will have the highest continuous-wave transmitter output available, one megawatt (1,000 kilowatts).

Radar sensitivity relates to the ability to receive signals. One of the limits to radar sensitivity is the noise created by the antenna and receiver systems, not to mention cosmic background and extraneous terrestrial radiation, all of which is expressed as noise "temperature" in Kelvins (abbreviated K), analogous to the temperature scale of the same name. The higher the system temperature in Kelvins, the noisier the radar and the lower its sensitivity. In 1958, the Millstone radar had an overall system temperature of 170 Kelvins, while the more sensitive JPL radar receiver had an overall system temperature of 64 K in 1961. Although impressively low in their day, these temperatures today are judged intolerably high.

[278] Planetary radar astronomy borrows much of its terminology from optical astronomy, although not always retaining the original meaning. Facilities for conducting radar astronomy research are called observatories and the instruments telescopes. Radar telescopes "illuminate" the surface of targets. The reflecting geometries of radar telescopes, called their "optics," take their names from optical instruments (Cassegrainian and Gregorian subreflectors, for example).

Detection and ranging are two of the elemental observations made by planetary radar astronomers. A detection occurs when a radar antenna transmits waves toward a suspected target, the target reflects those waves, and a radar antenna receives the reflected waves (echoes) from the target. If we were crossing the Atlantic Ocean aboard a fictional ship, say the U.S.S. Marconi, we could use a radio transmitter and receiver, acting as a simple radar system, to detect the presence of icebergs or other ships. The radio pioneer Guglielmo Marconi suggested doing precisely that in a speech delivered in 1922.

If our U.S.S. Marconi radar were to detect the presence of another ship, we could determine the distance from the Marconi to the other ship with our radar. Measurements of the distance to the target are called range, time-delay, or delay measurements. The ability to use radar to measure range is based on the knowledge that radio waves travel at a constant speed, namely, the same speed as light.

In order to determine how far away a target is, we simply measure how long it takes the echoes to arrive at the receiver antenna. The greater the distance to the target, the longer the echoes take to appear in the receiver. Conversely, the shorter the distance to the target, the less time the echo takes to appear in the receiver. The time between the moment of transmission and the moment the echo is detected can vary considerably. For the farthest bodies detected by radar astronomers, such as Saturn's rings, the signal round-trip travel time is about two and a half hours, while the round-trip travel time to some asteroids detected close to Earth is about two and a half seconds.

Another basic planetary radar measurement is Doppler frequency shift. Whereas range measurements indicate the distance between the radar observer and the target, Doppler shift indicates the motion of the target relative to the observer. With our fictional U.S.S. Marconi radar, we can determine not only the presence and distance of another ship, but its speed toward or away from us as well.

Radar transmitters send waves at a specific frequency. A perfect reflection from a motionless target appears at the radar receiver (after Fourier transformation, see below) as an almost line-like peak. The echo from an actual solar system target, however, is spread over a range of frequencies. This frequency spread is called a spectrum (plural spectra). In radar astronomy experiments, the motions of the Earth, on which the planetary radar sits, and the motions of the target are far more complex. The Earth spins on its axis and rotates around the Sun, while the target planet similarly spins and rotates. The relative motions of the Earth and target planet cause what is known as the Doppler effect or Doppler shift (or even Doppler offset).

Simply stated, the Doppler effect causes the frequency of radar echoes to differ from the transmitted frequency. The Doppler effect on sound waves is a rather common experience around high-speed transport. If we stand alongside railroad tracks, or a freeway, we can detect the Doppler effect with our ears. As a train rapidly approaches, the sound of the train seems to rise in frequency, that is, in pitch; as the train travels away from us, its sound seems to fall in frequency. The same Doppler effect occurs in radar. Depending on the line-of-sight motion of a planet (or whether the object is approaching or moving away from the observing radar), the frequency of the planet's echo will be higher or lower than the transmitted frequency.

[279] Planetary radar astronomers want to remove the average Doppler effect in order to analyze the information contained in the Doppler spectrum spread, so they use a radar ephemeris program. Although the average Doppler effect sometimes is removed in the transmitter, such as when several antennas receive, in general it is removed in the receiver. An ephemeris (plural ephemerides) is an astronomical term that refers to a set of tables that indicate the position of a planet or other body in the sky. A radar ephemeris program is computer software linked to the radar receiver that automatically adjusts the incoming signal for the expected Doppler shift. The amount of Doppler shift predicted by the ephemeris program must be accurate enough to avoid smearing the echo in frequency, and this requirement places stringent demands on the quality of the radar ephemeris.

Once we know the range and Doppler shift values of a solar system target, we can construct a two-dimensional radar image of the target called a range-Doppler or delay-Doppler image. Maps made from these images are vital to the exploration of the solar system, especially the planet Venus, whose surface is obscured by clouds.

In range-Doppler imaging, we assume that the target is a perfect sphere. An exception is the case of asteroids, whose nonspherical shapes require special modeling techniques. The transmitted radar waves arrive first at the area on the planet's surface that is closest to Earth. This area of initial impact is circular, because we have assumed that the target is spherical. The point on the planet's surface that is closest to the observer is called the subradar point. If we could look at the target with radar sensitive eyes, we would see a relatively dark sphere with a small bright spot in the middle, rather like a shiny ballbearing being held up to the light.


Figure 42. Diagram showing intersection of Range rings and Doppler strips to form a planetary range-Doppler image.

Figure 42. Diagram showing intersection of Range rings and Doppler strips to form a planetary range-Doppler image. The lines of constant phase permit resolution of north-south ambiguity. (Courtesy of Alan E. E. Rogers.)


[280] Echoes from the area beyond the subradar point are fewer than those that produce the central bright spot. Moreover, they reach those areas later than the waves striking the subradar point, because they have a greater distance to travel. The range values for those areas toward the limbs, then, are greater than those for the subradar region. Looking again at the target with our radar sensitive eyes, we see that the areas at a constant distance (range) from the radar transmitter form rings around the subradar point. These are called range rings.

As the planet spins on its axis toward or away from the oncoming radar waves, the spinning motion creates a Doppler effect. The Doppler frequency shift is the same along a strip or slice running across the planet's surface, because within each Doppler strip of the planet's surface, the motion relative to the observer is the same. When range and Doppler measurements made at the same time are combined, the strips of equal Doppler shift intersect the range rings to form "cells." Each range-frequency cell (or resolution cell) corresponds to a particular area on the planet's surface. The amount of area in a particular cell represents the amount of resolution of the radar image.

In range-Doppler imaging, any given range ring passes through the same Doppler strip at two points. One point is in the northern hemisphere, the other in the southern hemisphere of the planet. The two points have the same range and Doppler values, because they are in the same range ring and the same Doppler strip. As a result, these two points are indistinguishable in the radar image although they are in different hemispheres. Radar astronomers call this problem north-south ambiguity.

We can resolve the north-south ambiguity on the Moon by using a radar whose beamwidth is narrower than the diameter of the Moon. The beamwidth is the area of sky subtended by the radar beam. Astronomers measure beamwidth, and all solar system objects, in minutes and seconds of arc. The diameter of the Moon is a half degree or 30 minutes of arc. With a beamwidth of only 10 minutes, we can aim the radar antenna so that the subradar point is 10 minutes of arc north of the lunar equator. Echoes are not received from most of the southern hemisphere, so that echoes from the two hemispheres do not overlap. However, the technique is applicable to only the Moon. Compared to the Moon's 30 minutes of arc, Venus is only a speck; its diameter is but one minute of arc. Asteroids are less then one second of arc across.

In order to resolve north-south ambiguity on planetary targets, radar astronomers sometimes use a technique called interferometry. An optical interferometer is an instrument for analyzing the light spectrum by studying patterns of interference, that is, how light waves interact with each other. Radio astronomers began designing interferometers in the late 1950s. Planetary radar interferometry derived directly from those interferometers. Radio interferometers use two or more radio telescope antennas arranged along a line (called the base line). The separate antennas are linked electronically, so that the signals received at different points along the base line can be combined, compared, and studied with elaborate computer programs. Radar interferometers are somewhat simpler.

A radar interferometer consists of two antennas. The primary antenna transmits signals to the target and receives them. A secondary antenna, located not far (say, 1 to 10 km distant) from the primary antenna, also receives the echoes. Although a three-antenna radar interferometer was attempted between 1977 and 1988,2 in practice radar interferometers use only two antennas.

[281] The echoes received by both antennas are fed into a complex computer program that combines the echoes and obtains the fringe size (amplitude) and phase for each range-Doppler resolution cell. The computer program rotates the fringe pattern so that the lines of constant phase are perpendicular to the strips of equal Doppler value. The north-south ambiguity now is resolved, because the phase at points A and B have distinct phase values.

During the 1950s, researchers underwritten by the military developed a similar radar imaging process that used both range and Doppler data. However, that process involved imaging the Earth from aircraft and relied on developing a radar "history" of the target to create an image, while planetary range-Doppler mapping created a "snapshot" of a planetary surface from a ground-based radar. The airborne imaging process, called synthetic aperture radar, has since played a key role in the mapping of Venus by the Magellan spacecraft.

Radar astronomers do not depend entirely on range and Doppler data, however. The echo from a solar system target exhibits a number of attributes. From their analysis of those attributes, radar astronomers draw conclusions about the characteristics of the target. For example, the shape of power spectra can provide information about a target. If we aim at an asteroid and get an echo with two major peaks, called a bimodal echo, we can interpret the echo as possibly indicating a bifurcated shape, perhaps two asteroids joined together. Radar observations of asteroid 4769 Castalia (1989 PB), for instance, revealed it to be a contact binary asteroid.3

Small detail features on power spectra also can reveal vital information about a target's motion. For example, radar observations of Venus made in 1964 indicated that planet's rotational rate and direction. The radar instrument was both sufficiently powerful and sensitive that a large feature on the planet's surface showed up in the power spectra as an irregularity or "detail." The detail resulted from the fact that the surface feature scattered back to the radar antenna more energy than the surrounding area.

On close examination, one irregularity in the power spectra persisted day after day and appeared to change its position slowly. A study of the irregularity's movement led to a calculated rotational rate for the planet, but not immediately its prograde (forward) or retrograde motion. That information came from measurements of the width of the lower portion of the power spectra. Those widths were compatible with only a retrograde motion.

In the normal, round-trip journey of a radar wave from transmitter to target to receiver, a certain amount of power is lost. The amount of that loss is given by the so-called radar equation. The amount of power that reaches a target is inversely proportional to the square of the distance to the target, but the amount of power returned from the target to the receiving antenna also varies inversely proportional to the square of the distance to the target. After the complete round-trip from transmitter to receiver, the amount of power that arrives at the receiving antenna varies inversely with the distance to the target raised to the fourth power, that is, the square of the square of the distance. The radar equation shows that large amounts of power (hundreds of kilowatts) must be radiated into space in a very narrow beam in order to detect a target.

The amount of power returned from a target can reveal much about its surface characteristics. The total power returned from a target is a function of its radar cross section [282] or backscattering coefficient, that is, the target's ability to reflect energy to the radar receiving antenna. Radar astronomers express the radar cross section of a target in terms of an equivalent, perfectly reflecting surface. If a target scatters power equally in all directions, its cross section is equal to the geometric area of the target. That is the case of our ideal ballbearing target. For a perfectly reflecting spherical target, the radar cross section is one fourth the total surface. Surface irregularities affect the amount of power returned (or scattered back) from a target. Radar echoes have two scattering components, called quasispecular and diffuse. The quasispecular component arises from mirror-like reflections from parts of a flat or gently undulating surface. Those surface facets are perpendicular to the line of propagation, so they direct a large amount of energy back toward the observer. Such echoes concentrate at the center of the planet's visible disk, that is, around the subradar point, because the likelihood of finding favorably oriented facets is highest where the surface is perpendicular to the incoming radar beam. The diffuse scattering component comes from objects and structures with irregular shapes and therefore facets that redirect much of the radar beam away from the observer. The signal returned from the areas toward the limbs is called diffuse.

The amount of power returned from a target is, therefore, a consequence of the scattering component, quasispecular or diffuse, and the angle of a surface facet relative to the line of propagation of the radar wave. Flat surfaces perpendicular to the line of propagation return power directly back to the radar. The greatest amount of power returned from a target, then, comes from flat surfaces that are perpendicular to the line of propagation. If the reflecting surface is not perpendicular to the line of propagation, then power will be reflected away from the radar, and the amount of power returned to the radar antenna will diminish. The reduction in power will increase as the surface is tilted away from the radar.

For example, if a planetary target has mountains or craters, a portion of the radar power will be reflected away from the return path, depending on the angle, that is, the amount of slope, of the mountain or crater. The more power returned, the gentler is the slope or angle of the surface. Factors other than surface slope can affect the amount of power returned from a target, too.

If a planetary surface is covered by boulders or other material with multiple sides, a complex scattering process takes place. Some power is returned to the radar, some power is deflected away from the radar return path, while some power scatters among the boulders. If the surface is covered by material significantly smaller than boulders, say volcanic ash, the loss of power from scattering in directions other than the return path can be considerable.

Although range-Doppler mapping techniques provide one means for correlating echo power spectra and surface features, they are not always practical. Other methods must be used. For example, Mars rotates much faster than Venus, whose slow retrograde motion makes it an ideal radar target. Mars is what radar astronomers call an overspread target. The rapid rotation of that planet means that the signal from one range ring spreads over into the next ring, or the signal from one Doppler strip spreads over into the next strip. Also, the echo from Mars is much weaker, because the distance to Mars is greater than to Venus.

In helping to select a landing site for the Viking lander, for example, radar astronomers relied on a different approach to interpret the amount of power returned from an area of the surface. In this approach, a geometric model for the entire visible surface of the planet was assumed. These models, or scattering laws as they are called, also can be derived empirically from actual radar observations of the target, if the target surface is sufficiently well known. The most commonly used model is the Hagfors scattering [283] law, named for its originator, Tor Hagfors, an ionosphericist and radar astronomer who is currently Director of the Max Planck Institut für Aeronomie.

In studying candidate sites for the Viking lander, radar astronomers at Haystack, Arecibo, and Goldstone applied the Hagfors scattering law and determined the slope, that is, the degree of tilt from the line of propagation, of various areas of the Martian surface. Site candidates had to be inclined no more than 19 degrees; otherwise, the lander would topple over. They also had to be free of rocks and other objects larger than 22 cm, the height of the lander vehicle.

Radar data were capable of indicating the roughness of the Martian surface down to a few centimeters, while photographs had a resolution of roughly 100 meters, larger than a football field. The radar data, however, was not expressed visually, like the photographs, but mathematically as the root-mean-square (rms) slope. The rms is a special mathematical method for averaging. The rms slope gives an indication of the average slope or inclination of a given area of the Martian surface. When applying the Hagfors scattering law, the value for the rms slope varies in theory up to three degrees, the upper limit for the validity of the assumptions underlying the model. In practice, however, the Hagfors scattering law yields much higher values of rms slope.

One of the most important signal parameters used in planetary radar astronomy today is polarization. The earliest lunar radar experiments carried out at Jodrell Bank used a linear antenna feed. Antenna feeds are either linear or circular, and the feed shape determines the polarization of the transmitted wave. When Jodrell Bank investigators sent radar signals to the Moon, they discovered two patterns of signal fading. Normal lunar libration caused the slowly fading echoes, while the rapidly fading signals, they concluded, resulted from the radar waves passing through the Earth's ionosphere.

In the 19th century, the British scientist Michael Faraday discovered that a magnetic field could alter the plane of polarization. The effect since has come to be known as Faraday rotation. In the case of the lunar radar signals, the Earth's magnetic field rotated the signals' plane of polarization as they passed through the ionosphere.

A radar target also can change the handedness, or rotational sense, of circularly polarized waves. If we transmit circular waves with right-handed polarization to a perfectly reflecting target, the power returns with a left-handed polarization. If we transmit right-handed polarization and adjust the antenna feed to accept right-handed polarization, the antenna will detect little or no returned power. In practice, when dealing with the inner planets (Mercury, Venus, and Mars), most power returns from a planetary target in the opposite sense of polarization in which it is transmitted.

Exceptions to this rule arose when radar astronomers began exploring the Galilean moons of Jupiter and other icy targets. When radar signals return from the surfaces of those targets, somewhat more of the power is received in the same sense of polarization. In other words, if we transmit right-hand polarization, more of the power will return with right-handed polarization. The peculiar nature of these radar targets has elevated the importance of the ratio of same sense to opposite sense polarization as a radar measurement. Radar astronomers now transmit one sense of polarization and receive both right-hand and left-hand polarization, then they compare the right-hand and left-hand values. While fractured ice is the apparent cause of this polarization phenomenon, the mechanism that gives rise to it is only now beginning to be understood.

Planetary range-Doppler radar experiments take place in several stages: data taking, decoding, rotating the matrix, the Fourier transform, conversion into latitudes and longitudes to create maps. The first stages, especially data taking and decoding, are rather routine and standardized; software specialization tends to take place in the last stages of the process, such as converting the data into latitudes and longitudes.

[284] While data taking involves the combined use of the radar system and an associated computer, the remaining stages take place entirely on a computer. Computer time accounts for most of the processing time spent on a range-Doppler experiment. With modern computer technologies that accelerate processing time, data reduction takes far less time than before. At the Arecibo telescope, a typical run of observations on Mercury takes about 10 minutes. Data processing of those 10 minutes of radar activity consumes another hour and a quarter to an hour and a half of computer time. Roughly, then, every minute spent making radar observations translates into eight minutes of processing time.

Without those special accelerating technologies, a mainframe computer takes far more time; it is almost 80 times slower. One run, then, might take an entire day to process. Older mainframe computers were even slower. The competition for computer time was often as intense as it was for antenna time. Moreover, these computer times apply only to data reduction, the initial preparation of the data. Analysis, modeling, and interpretation can be far more time consuming.

The first stage of a planetary radar experiment is the recording of the raw echoes as they come from the antenna through the receiver. In the earliest lunar radar experiments conducted at Jodrell Bank in the 1950s, the echoes were observed on an oscilloscope and recorded with a cinema camera. The films are extant and form part of the archives deposited with the University of Manchester. Beginning with the first attempts on Venus in 1958, the raw signals were recorded on magnetic tape. In addition, they were routinely converted from analog into digital signals for processing. Today, all planetary radar astronomy is carried out digitally.

The unprocessed echoes are usually too weak and too noisy to process, so the echoes are accumulated together. The next stage is to decode the signals. Before transmission, the signal is encoded with a repeating binary code. Pulse radars achieve binary coding by turning the signal off and on. With continuous-wave radars, binary coding is accomplished by changing the phase of the signal. These off-on states and phase changes in the coded transmission tell the radar astronomer which part of the wave is being examined. An accompanying time code identifies the location on the planet where the particular echoes originate.

The next step in forming a planetary radar map is to rotate the matrix. Once the codes have been removed from the echoes, the signals are arranged in a matrix that corresponds to the various range rings on the planet. The computer software looks at each code cycle and considers each range ring separately. The values for a given range ring are a function of time. The software now must decide which frequencies are present, in order to find the Doppler delay values.

A Fourier transform sorts echoes from a given range ring into frequency bins. A Fourier transform is a specific type of transform, a powerful mathematical expression that transforms (hence the name) one geometrical figure or analytical expression into another. During the 1960s, the Fast Fourier transform was devised by engineers in the field of signal processing. As a result, a mathematical operation that previously took 30 minutes on an IBM 8094 mainframe computer now took only about five seconds.4

The result of the Fourier transform is a range-Doppler, two-dimensional picture. Additional analysis with various computer algorithms written in the software can yield a three-dimensional picture. However, the three-dimensional picture requires adding further information to the range-Doppler map.

This succinct cursory overview of a planetary radar experiment is limited to only range-Doppler mapping. Radar astronomers carry out several other types of experiments. Most experiments are routine and rely on cookbook software and processing. Radar [285] interferometry and the random code technique, adapted from ionospheric radar research, for example, employ specialized software and processing techniques.

A number of articles and book chapters on radar astronomy published since 1960 discuss the field, its accomplishments, and its techniques. Their intended audience runs the gamut from general to specialized. They are recommended to those wishing information on radar astronomy beyond that provided here.



1. Louis A. Gebhard, Evolution of Naval Radio-Electronics and Contributions of the Naval Research Laboratory, Report 8300 (Washington: NRL, 1979), p. 170.

2. Jurgens, Goldstein, Rumsey, and R. Green, "Images of Venus by Three-Station Radar Interferometry: 1977 Results," Journal of Geophysical Research 85 (1980): 8282-8294.

3. Hudson and Ostro, "Shape of Asteroid 4769 Castalia (1989 PB) from Inversion of Radar Images," Science 263 (1994): 940-943.

4. Gwilym M. Jenkins and Donald G. Watts, Spectral Analysis and its Applications (San Francisco: Holden-day, 1968), pp. 313-314.


For Further Reading


Eshleman, Von R. "Radar Astronomy: Exploration of the Solar System Using Man-made Radio Waves." In Aeronautics and Astronautics: Proceedings of the Durand Centennial Conference, Stanford, August 5, 1959, edited by N. J. Hoff and W. G. Vincenti, 207-226. London: Pergamon Press, 1960.

Eshleman, Von R and Alan M. Paterson. "Radar Astronomy." Scientific American 203 (1960): 50-54.

Evans, John V. "Radar Astronomy." Contemporary Physics 2 (1960): 116-142.

Evans, John V. "Radar Astronomy." Science 158 (1967): 585-597.

Green, Paul E., Jr., and Gordon H. Pettengill. "Exploring the Solar System by Radar. Sky and Telescope 20 (1960): 9-14.

Hagfors, Tor, and Donald B. Campbell. "Mapping of Planetary Surfaces by Radar. Proceedings of the IEEE 61 (1973): 1219-1225.

Jurgens, Raymond F. "Earth-based Radar Studies of Planetary Surfaces and Atmospheres." IEEE Transactions on Geoscience and Remote Sensing GE-20 (1982): 293-305.

Kippenhahn, Rudolf. Bound to the Sun: The Story of Planets, Moons, and Comets, trans. Storm Dunlop. New York: W. H. Freeman and Company, 1990, pp. 259-272.

Muhleman, Duane O., Richard M. Goldstein, and Roland Carpenter. "A Review of Radar Astronomy, Parts 1 and 2." IEEE Spectrum 2 (1965): 44-55 & 78-89.

Ostro, Steven J. "Planetary Radar Astronomy." Reviews of Geophysics and Space Physics 21 (1983): 186-196.

Ostro, Steven J. "Planetary Radar Astronomy." In Encyclopedia of Physical Science and Technology, edited by Robert A. Meyers, 10:611-634. Orlando: Academic Press, 1987.

Ostro, Steven J. "Planetary Radar Astronomy." Reviews of Modern Physics 65 (1993): 1235-1279.

Ostro, Steven J. "Radar Astronomy." In McGraw-Hill Encyclopedia of Astronomy, edited by Sybil P. Parker and Jay M. Pasachoff, 347-348. New York: McGraw-Hill, 1993.

[286] Pettengill, Gordon H. "Radar Astronomy." Transactions of the American Geophysical Union 44 (1963): 453-455.

Pettengill, Gordon H. "Planetary Radar Astronomy." In Solar System Radio Astronomy, edited by Jules Aarons, 401-411. New York: Plenum Press, 1965.

Pettengill, Gordon H. and Irwin I. Shapiro. "Radar Astronomy." Annual Review of Astronomy and Astrophysics 3 (1965): 377-410.

Pettengill, Gordon H. "Radar Astronomy." International Science and Technology 58 (1966): 72-74, 76, 78, 80-82.

Shapiro, Irwin I. "Planetary Radar Astronomy." IEEE Spectrum 5 (1968): 70-79.

Thomson, John H. "Planetary Radar." Quarterly Journal of the Royal Astronomical Society 4 (1963): 347-375.

Thomson, John H. "Planetary Radar." Science Progress 53 (1965): 183-190.