SP-4218 To See the Unseen


- Chapter Six -

Pioneering on Venus and Mars



[149] Range-Doppler mapping and radar techniques for determining the roughness, height variations, and other characteristics of planetary surfaces came into their own in the early 1970s and shaped the kinds of problems planetary radar could solve. Radar techniques and the kinds of problems they solved were cross-fertilizing forces in the evolution of planetary radar astronomy. In the early 1970s, NASA was shifting gears. The landing of an American on the Moon, the zenith of the Apollo program, was history when in December 1972 Apollo 17 became the last to touch down on the Moon. Now the unmanned exploration of the planets began in earnest.

The usefulness of radar to planetary exploration had been argued by radar astronomers as early as the 1959 Endicott House conference. However, not everyone shared their enthusiasm. Smith and Carr, for example, in their 1964 book on radio astronomy, wrote: "It is inevitable that the importance of the exploration of the planetary system by radar will diminish as instruments and men are carried directly to the scene by space vehicles. However, that time is still to come. In the meantime, the information that radar provides will be vital in man's great effort to conquer space."1 Soviet radar astronomers B. I. Kuznetsov and I. V. Lishin expressed similar sentiments in 1967: "Certainly, radar bombardment of the planets gives less information than a direct investigation of them with spaceships and interplanetary automatic stations." However, they did foresee that information about planetary surfaces would "help designers in the development of spaceships intended for making a 'soft' landing on the planets."2

As NASA came to fund planetary radar research, experiments and NASA missions became linked. Goldstone antenna time depended on mission approval, while Haystack radar funding was tied to specific, mission-oriented tasks. It is not surprising, then, that planetary radar in the 1970s evolved in point and counterpoint to the NASA space program, at first modestly to correct data returned from Soviet and American missions to Venus, next to help select a Mars landing site, and then to image Venus from a spacecraft. This evolution followed from the precedent established by NASA's funding of lunar radar imaging for the Apollo program. The Pioneer Venus radar imaging and altimetry missions took radar astronomy off the ground and into space. Again, just as ground-based radar astronomy had piggybacked itself onto Big Science radio astronomy facilities, so the Pioneer Venus radar attached itself to a larger mission to explore the planet's atmosphere.

The new techniques and problem-solving activities drew radar astronomers into closer contact with planetary scientists from a variety of disciplines who were not necessarily familiar with radar or the interpretation of radar results. It was one thing for radar astronomers to determine a spin rate for a planet or the value of the astronomical unit; astronomers easily grasped those discoveries. However, when radar astronomers described planetary surfaces in such abstract terms as root-mean-square slope to geologists, whose discipline rests heavily on hands-on field knowledge, a communication problem arose and serious misinterpretations and misunderstandings of radar results ensued.


[150] The Radar Radius of Venus


On 18 October 1967, the Soviet Venera 4 space probe entered the atmosphere of Venus and began to transmit data back to Earth. From that data, Soviet scientists calculated a value for the radius of Venus, 6,079 ± 3 km, on the assumption that the break in the probe's transmissions indicated that it had reached the planet's surface. On the following day, Mariner 5 passed within 4,100 km of Venus and conducted a series of experiments. From the data beamed back to Earth, Mariner scientists at JPL calculated a value for the radius of Venus that was compatible with that determined by their Soviet colleagues, 6,080 ± 10 km.

The data from Venera 4 and Mariner 5 were consistent with each other and with the latest optical data, which yielded a value of 6,089 ± 6 km. However, the space and optical values differed markedly from the size of the radius, 6,056 ± 1.2 km, determined by Irwin Shapiro, Bill Smith, and Michael Ash with the Lincoln Laboratory radars as part of the Planetary Ephemeris Program.3

If the spacecraft and optical measurements were correct, then the radar data or its analysis were in error. The radius of Venus was a critical radar measurement; its value, for example, could serve to study the planet's topography. Radar astronomers associated with MIT and the Haystack Observatory, Gordon Pettengill, Irwin Shapiro, Dick Ingalls, Michael Ash, and Marty Slade, and those at the Arecibo Observatory, Rolf Dyce, Don Campbell, Ray Jurgens, and Tommy Thompson, took up the challenge in collectively authored papers that appeared in Science and the Journal of the Atmospheric Sciences. The publications embraced both a general audience and atmospheric specialists.

In addition to data collected previously at Millstone, Haystack, and Arecibo, the MIT-Arecibo radar astronomers added data from fresh radar observations made in 1966 and 1967 as well as optical observations from the U.S. Naval Observatory from the period 1950 through 1965. The magnitude of the data base was impressive and convincing. The Arecibo and MIT investigators analyzed their data separately and obtained radii of 6,052 ± 2 km and 6,048 ± 1 km, respectively. They concluded that Mariner 5 had misjudged its distance from the planet's center by about 10 km, and that "the simple possibility that Venera 4 underestimated its altitude by about 35 km cannot yet be ruled out."4

Dewey Muhleman, now professor of planetary science at the California Institute of Technology, with Bill Melbourne and D. A. O'Handley of JPL, made observations of Venus between May 1964 and October 1967 with the Goldstone Mars Station. Because their data were reported only in internal JPL reports, Lincoln Laboratory did not use that data. Consequently, they asserted, their observations constituted "an entirely independent data source." Muhleman and his JPL colleagues determined a value for the radius of Venus of 6,053.7 ± 2.2 km, in strong agreement with the MIT and Arecibo results.5

Arvydas Kliore and Dan L. Cain, two JPL scientists on the Mariner mission, saw the agreement between the Caltech-JPL and the Arecibo-MIT values and realized that "the consistency between reductions from data taken by different radars and reduced by different investigators cannot be ignored." They discovered that the different timing systems [151] used by the Deep Space Network to acquire Mariner 5 data, namely Station Time and Ephemeris Time, had introduced an error into their calculations. The amount of that error, 8.85 km, brought the Mariner 5 value for the radius of Venus in line with the radar results.

To explain what was now the anomalous Soviet value for the radius of Venus, Kliore and Cain concluded that either the Venera 4 capsule landed on a peak or plateau that was about 25 km high and not detected by planetary radar or the capsule stopped transmitting before reaching the solid surface of Venus. The problem with Venera 4, Don Campbell ventured, "was tied up in an ambiguity difficulty in their own radar system, which was a pulsed altimeter radar. I think, frankly, that the scientists who reported the results did not know how it worked. It was a military radar altimeter. They were just provided the answer, essentially. Although I don't know, and probably didn't know at the time either, what exactly the circumstances were, that was the impression that one got."6


"A Little Radar Knowledge is a Dangerous Thing."


Well before radar astronomers began collaborating with geologists, misinterpretations of radar data occurred. In fact, radar astronomers themselves were not immune to misconstruing radar results, as the case of the radar brightness of Mars illustrates. In initial observations of that planet, radar astronomer Dick Goldstein assumed a relationship between radar brightness and optical darkness. Arecibo observations appeared to confirm that relationship, which snowballed among planetary astronomers into a hypothesis that correlated radar brightness and topography (continental blocks and dry ocean basins). A reconsideration of evidence showed no such correlation.

When Dick Goldstein made his pioneering radar observations of Mars in 1963, he discovered what he thought was a relationship between radar "brightness," that is, the average amount of power returned in the echo from a given surface area of the planet, and the optical darkness of that same surface area. Goldstein constructed what he called a radar map of Mars, which showed variations in radar brightness. He noted, for example, that the Syrtis Major region appeared bright to the radar, but dark to visual observations. Because radar brightness is a function of surface roughness, he argued, the brightest radar areas were regions of flatness, while dark radar areas were topographically rough.7

In 1965, Goldstein observed Mars at the next opposition and again looked at the radar brightness of the planet's surface, this time at latitude 21° North. The average power returned (radar brightness) reached a maximum in the region of Trivium Charontis (an optically dark area), then dropped off abruptly when the neighboring area of Elysium (optically bright) was the radar target. Based on the known relationship between surface roughness and radar brightness, Goldstein concluded the existence of a very smooth, strongly reflecting area extending 20° to 30° in longitude and having an unknown latitudinal extent in the region of Trivium Charontis.8

During the same opposition, Gordon Pettengill, Rolf Dyce, and Don Campbell observed Mars with the UHF radar at Arecibo. When they compared their results with an optical map of Mars, the Arecibo investigators found a general tendency for weak echoes to correlate with the (optically) lighter areas of Mars, such as Arabia, Elysium, Tharsis, and [152] Amazonis, and a tendency for strong echoes to correspond with visually darker features, such as the regions near Trivium Charontis and Syrtis Major. They did, however, note that the correlation between radar brightness and optical lightness was not perfect. For instance, the peak radar echo near Trivium Charontis occurred at 201° longitude, which is on one edge of the visually dark region. Likewise, the visually darkest region of Syrtis Major corresponded to a local minimum in echo strength.9

The Arecibo results were rather convincing. Not only had they been obtained from roughly the same area (22° North latitude) that Goldstein had studied, but the Arecibo and Goldstone observations had been made at two different frequencies (UHF vs. S-band). The persistence of the correlation between optical darkness and radar brightness at both frequencies was persuasive.

Astronomers Carl Sagan and James B. Pollack, then at the Smithsonian Astrophysical Observatory, and Richard Goldstein carried out a lengthy and detailed analysis of the JPL 1963 and 1965 radar data. They maintained and extensively documented the correlation between high radar reflectivity and optical darkness, despite some exceptions. Not only did radar bright and optically dark areas correlate; they claimed that topography and radar brightness also were related. Dark areas were elevations similar to continental blocks; bright areas were comparable to dry ocean basins.10 The notion that Martian dark areas were elevated land masses rapidly gathered support from other planetary astronomers in the United States and Britain.11

Nonetheless, Pettengill, who had participated in the earlier effort at Arecibo, now opposed the correlation of visual darkness and radar brightness and undertook observations at Haystack, during the 1967 opposition, specifically in order to oppose the prevailing hypothesis that now correlated topography and radar brightness. Pettengill conducted a series of straightforward, precise range measurements to establish the topographical variations along latitude 22° North. Then he compared those range measurements with the average planetary radius taken from the planetary ephemeris data. He also plotted echo power over longitude along that same latitude.

Pettengill found no significant correlation between radar brightness and topography. A direct comparison between the radar results and a map of visible Martian surface features revealed no clear one-to-one association between bright or dark areas and topographical extremes. What others had observed as variations in radar brightness, Pettengill argued, resulted from the deviant properties of relatively small regions of the surface near the subradar point. Moreover, he pointed out, arguments for the hypothetical correlation between elevation extremes and brightness had been based largely on conclusions drawn from a range of disparate isolated locations. Further Haystack observations of Mars carried out under Pettengill's direction reinforced the conclusion that no correlation existed between regions of high radar reflectivity and optically dark areas.12

Perhaps one of the most notorious examples of misinterpreted radar results is that of Thomas Gold of Cornell University. Gold had been developing theories about the lunar surface since the 1950s. Long before he ever saw any radar data, Gold favored a meteoric [153] explanation for lunar craters and developed an explanation for the presence of vast flat level surfaces that did not require the deposition of volcanic lava. His hypothesis was that these flat expanses consisted of dust from meteoric impacts. Gold interpreted radar observations of the Moon as supporting the existence of a surface layer of fine rock powder several meters deep, which a seismic experiment carried out by Apollo 12 allegedly supported. The implications for landing an American on the Moon were obvious; an astronaut might sink several centimeters into the powder or even "wallow" in it.13

Many scientists greeted Gold's prediction of a deep layer of powder with disbelief. As Don E. Wilhelms wrote, "Four Surveyor and six Apollo landings established the strength, thickness, block content, impact origin, and paucity of meteoric material in the Moon's regolith. There is fine pulverized soil, but it is weak only for a few centimeters of its thickness. Yet Thomas Gold is still fighting the battle. Still believing radar more than geological sampling..."14 Wilhelms went so far as to state, "A little radar knowledge is a dangerous thing."15 Gold later defended himself by insisting that although the "Gold dust" (as it has come to be called) would be many meters thick, the idea of sinking in it was a "total misconception."16

The Apollo program started the process of bringing together radar astronomers and geologists. The lunar radar images created by Tommy Thompson and Stan Zisk from data gathered at Arecibo and Haystack contributed not inconsequentially to America's exploration of the Moon. On occasion, nonetheless, radar astronomers misinterpreted lunar landing sites. In one instance, a landslide was mistaken for a field of boulders at the Apollo 17 landing site, while in another radar astronomers incorrectly characterized the roughness of the Apollo 14 Cone Crater site. These problems, however, arose not from mistaken readings of radar images, but from misinterpretations of the root-mean-square slope and dielectric constants of the surface.17


Landing on Mars


During the preparation for the Viking mission to Mars, radar astronomers encountered the challenge of making radar data understandable to NASA mission personnel unfamiliar with the interpretation of radar results. Until Congress funded the Voyager mission to Jupiter and Saturn, Viking was NASA's biggest and most expensive program for planetary exploration. Viking was to land on that planet, and NASA needed a landing site that was both safe for the lander and interesting to scientists. Radar astronomers collected and interpreted data to help with the selection of candidate sites.

The selection of the Viking lander site also brought together ground-based planetary radar astronomy and the Stanford bistatic radar approach under the aegis of NASA. Ground-based planetary radar astronomy had distinguished itself from "space exploration" (the Stanford approach), but the boundary between ground-based planetary radar astronomy and "space exploration" softened, as radar astronomers played an expanding role in NASA missions of planetary exploration and as Stanford investigators extended their field of applications.

[154] Images of Mars from earlier missions provided a clue in selecting candidate Viking landing sites. As early as 1965, Mariner 4 had flown past Mars and snapped 22 pictures of about one percent of the planet's surface. Mariner 6 and Mariner 7 took about 200 images of around 10 percent of the surface in 1969. The goal of Mariner 9, to make a complete photographic map of Mars was thwarted; when the spacecraft arrived at its destination, a planet-wide dust storm concealed most of the surface. Once the storm appeared to subside, Mariner 9 began to transmit images to Earth in early 1972, and the study of Martian topography began in earnest.18

Unlike the Mariner flybys, Viking was to study Mars by landing on its surface. A pair of orbiters was to focus on atmospheric studies, while a pair of landers studied the surface, if all went well. If the Viking landers were to touch down on a large rock or precariously on an edge, the entire mission might be lost. The clearance under the lander body was only 23 cm (nine inches), so a relatively smooth landing surface was a prime mission requisite.

NASA selected landing and backup sites for two landers. The sites had to be around 25° North latitude; at any other latitude, the orbiter solar panels would not receive sufficient solar energy to keep the spacecraft's batteries charged. That power was critical to the transmission of telemetry to Earth.

A major criteria for selecting candidate landing sites was the potential availability of water. Water meant the possibility of finding life, which was a major mission objective. Chryse, located at 19.5° North and 34° West, was scientifically interesting, because it is located at the lower end of a valley where the largest group of Martian channels diverges. The site may have been a drainage basin for a large portion of equatorial Mars and, therefore, would have collected deposits of a variety of surface materials.19

Despite the scientific interest in Chryse as the prime Viking landing site, the high-resolution Mariner 9 images lacked sufficient resolution to determine the site's safety. As Don Campbell recalled: "NASA was very concerned about how rough the surface was at the landing site. None of the Mariner 9 imagery had any hope of giving information at scales of 10 cm to a meter, which was the amount of surface roughness that they cared about."20 Mariner 9 images had a resolution of about 100 meters, roughly the size of a football field, and simply did not show objects small enough to jeopardize the touchdown of the lander, which had a clearance of only 23 cm. The radar data, in contrast, were capable of indicating surface roughness down to objects only a few centimeters across. Once again, radar was going to try to solve a problem left unresolved by optical methods.


The Stanford Center for Radar Astronomy


In order to help select candidate Viking landing sites, NASA turned to radar astronomy and its ability to appraise gross and fine surface characteristics. The chief advocate for the use of radar data was Carl Sagan. Sagan was concerned about the possibility that the first lander might disappear in quicksand at one of the equatorial sites. In general, he believed that too much stress had been placed on visual images with a resolution of only 100 meters and not enough on radar, which could indicate surface irregularities at the [155] 10-cm scale. Sagan urged further study of the meaning of the radar data, so that the properties of the Martian soil could be better evaluated.

In response to Sagan's urging, on 1 March 1973, Tom Young and Gerald Soffen, Viking science integration manager and project scientist, respectively, met with Von Eshleman and Len Tyler of the Stanford Center for Radar Astronomy. Both already were investigators on Viking with a radio scattering experiment. Young and Soffen asked Tyler to acquire, analyze, and interpret radar data and to set up a radar study team for the selection of Viking landing sites. Tyler agreed.21

The Viking Project Office probably approached the Stanford Center for Radar Astronomy because Eshleman and Tyler already were Viking investigators, but also because of the Center's experience in interpreting Doppler spectra from the lunar surface. The Stanford Center for Radar Astronomy (SCRA) was a joint venture of Stanford University and the Stanford Research Institute (SRI) created in 1962 to foster scientific and engineering efforts and to provide graduate student training in radar astronomy and space science. It was the umbrella organization for Eshleman and his program of bistatic radar astronomy. A NASA grant underwrote the Center itself, while additional military and civilian awards supported a range of theoretical and experimental radio and radar research on space, ionospheric, and communication theory topics.22

Len Tyler, as did his Stanford colleague Dick Simpson, brought considerable knowledge of radar techniques to the effort. A graduate of Georgia Institute of Technology, Tyler had been at the SCRA since 1967, when he received his doctorate in electrical engineering from Stanford under Von Eshleman. Tyler invited Dick Simpson to work on the Viking data. Simpson, a graduate of the MIT electrical engineering program, had joined the SCRA in 1967 as a research assistant while working on his MS and Ph.D. in electrical engineering.23

Later, during the 1978 Mars opposition, Simpson and Tyler conducted 29 bistatic radar observations using the Viking 1 and 2 orbiter spacecraft in conjunction with the DSN antennas at Goldstone and Tidbinbilla (near Canberra, Australia) to study Mars surface roughness and scattering properties, and Simpson made ground-based monostatic radar observations of Mars, not associated with the Viking project, at Arecibo.24 Their radar work, however, began much earlier, during the Apollo era.

For his doctoral thesis, Tyler had developed a method for creating two-dimensional surface images of the Moon using an Earth-based transmitter and a spacecraft receiver and based on theoretical work laid out earlier by another SCRA investigator, Gunnar Fjeldbo (now known as Lindal).25 Tyler first applied his bistatic imaging method on Explorers 33 (which missed the Moon) and 35, the first U.S. spacecraft to orbit the Moon, [156] and obtained crude meter-scale measurements of surface roughness and radar brightness.26 With Simpson, Tyler performed bistatic radar experiments on the Moon using the Apollo 14, 15 (at 13 and 116 cm), and 16 (at 13 cm only) command service modules while those vehicles were in lunar orbit; at the same time, they were receiving the S-band (13 cm) signals at Goldstone and the VHF (116 cm) signals at the Stanford 46-meter (150-foot) dish. However, Tyler and Simpson did not do imaging; they were more concerned with scattering mechanisms.27


Mars Radar


Tyler and Simpson began working on the Viking landing site selection problem by surveying and re-analyzing the available data. Radar data from several oppositions already were available, and those data obtained during the 1965 opposition were from the latitude of the preferred Viking landing sites, around 20° North. Radar studies of Mars made during the 1969 opposition provided useful topographical and surface roughness measurements, though not at latitudes interesting to the Viking mission. Haystack observed a swath of the planet's surface near the equator (latitudes 3° and 12° North), while Goldstone took observations at three latitudes (3°, 11°, and 12° North).28



Figure 25. Outline of Mars topography at 8° north of the equator released by JPL in July 1969.

Figure 25. Outline of Mars topography at 8° north of the equator released by JPL in July 1969. The outer white circle indicates a six-mile-high scale. The inner irregular line traces topographical variations found by radar. Syrtis Major and Trivium Charontis were found to be long slopes. The correlation of radar topographic data with known features in Mars photographic images aided geologists' ability to interpret the physical and historical geology of the planet. (Courtesy of Jet Propulsion Laboratory, photo no. 331-4539.)


The 1969 and earlier data, moreover, were too noisy to be of any use in sorting out a Viking landing site. The best data had been collected during the 1971 opposition, when Mars came closer to Earth than it would again for 17 years. Goldstone achieved its highest resolutions to date; results showed a rugged terrain, with elevation differences greater than 13 km from peak to valley. Altitude profiles showed heavy cratering, including several large craters 50 to 100 km in diameter and 1 to 2 km deep. Haystack also made high-resolution Mars radar observations during the 1971 opposition, measured surface heights with relative errors down to about 75 meters, and correlated craters detected by radar with those in images taken by Mariner.29

[158] Nonetheless, even that high-resolution data was not useful to the selection of a Viking landing site. Because of the geometries of the Earth and Mars during that opposition, planetary radar astronomers observed the southern hemisphere of the planet. The Goldstone radar observed Mars at latitude 16° South. Haystack observations during the 1971 opposition also examined southern latitudes.30 The best candidates for the Viking mission were in the northern hemisphere.

Thus, in 1973, when Tyler undertook the interpretation of Mars radar data for the selection of the Viking landing sites, radars had not observed the preferred Viking landing area near 20° North since 1967, nor any of the backup sites near the equator prior to 1975. The Viking Project Office funded a round of Mars radar observations in 1973 at the Haystack, Arecibo, and Goldstone radar telescopes at UHF, S-band, and X-band frequencies. Don Campbell and Rolf Dyce provided the Arecibo data, while Dick Goldstein and George Downs took the Goldstone data, and Gordon Pettengill furnished the Haystack data.31

The 1973 Haystack Mars data was placed in the same format as that obtained at Arecibo in order to facilitate their comparison. Although Haystack provided an abundance of radar data, its signal-to-noise ratio was generally too low for a detailed study of surface characteristics. The Haystack klystron was acting up,32 and Haystack ceased to participate in the Viking mission; shortly thereafter Haystack stopped all planetary radar experiments.

The 1973 Viking Mars data provided no direct information on potential landing sites. The orbital geometries of Earth and Mars meant that the subradar points of the three telescopes swept areas in the southern hemisphere, between latitudes 14° and 22° South far from either the main or backup landing sites.33 The 1973 data, nonetheless, provided an opportunity to better understand the radar properties of the Martian surface and for Tyler, in particular, to begin the difficult task of explaining the surface roughness of Mars in terms of root-mean-square (rms) slope to an audience unacquainted with the interpretation of radar data.


Radar Tutorials


Mariner images made the surface of Mars obvious to everyone. Radar data on surface roughness was not at all obvious and required expert interpretation. The "rift between believers in radar and believers in photography," in the words of Edward Clinton Ezell and Linda Neuman Ezell, first appeared at a meeting of the Viking landing site working group on 25 April 1972,34 well before Tyler and radar became a part of the site selection process.

[159] The key radar information on surface characteristics was not expressed visually, but mathematically. The abstract results were neither visual nor directly accessible by any of the senses. Moreover, the transformation of raw radar data into information on surface characteristics involved the interpretation of the data in terms of scattering laws and their expression as degrees of rms slope. The number of degrees of rms slope indirectly but reliably described the planet's surface roughness.

When a radar wave strikes an irregular planetary surface covered by boulders or other material with multiple sides, a complex scattering process takes place. Some power returns to the radar, some power is deflected away from the radar return path, while some power scatters among the boulders. The rougher the surface, the less power returns to the radar and the flatter is the return power spectra.

Because each radar target has a different surface makeup, its scattering behavior varies. Radar astronomers have sought general laws that describe scattering behavior. These scattering laws are mathematical descriptions of how much power is reflected back towards the radar at different angles of incidence. They are important tools for interpreting planetary radar data. At Haystack and Arecibo, radar investigators used what had become known as the "Hagfors Law," named after the Cornell University ionosphericist and radar astronomer.

The Hagfors Law mathematically expresses the general roughness of a planetary or lunar surface in terms of average slope. The root-mean-square is a specific type of mathematical average for the expression of these average slopes. When using the Hagfors Law, the value for the slope varies up to 3°, the upper theoretical limit for the validity of the assumptions underlying the Hagfors Law, although in practice much higher slope values are normal. The 1973 slope estimates for Mars ranged from 0.5° to at least 3°, suggesting that some areas, those closest to 0.5°, were suitable for a Viking landing. However, none of the 1973 radar experiments had observed areas of potential Viking landing sites.35

Tyler and the members of his radar study group presented their results to the landing site working group meeting at Langley on 4 November 1974. Tyler announced that his study group had learned a great deal: overall, the Martian surface was very heterogeneous; Mars tended to have greater variation in surface reflectivity than Earth or the Moon; and the planet appeared smoother than the Moon to the radar. However, he concluded, data acquired in the southern hemisphere could not be applied to northern latitudes without variation. Also, correlation between radar features and Mariner 9 imagery was poor.

Both Tyler and Gordon Pettengill "laced their presentations strongly with tutorial material which greatly enhanced the ability of the group to understand and correctly interpret their findings," reported Edward and Linda Ezell.36 After all, geologists would rather think about rocks than about Hagfors' Law, rms slopes, or dielectric constants, and those in charge of making the landing site had no knowledge of radar.37 The abstract nature of the radar data, as well as its complex and difficult interpretation, had an impact on the actual use of radar in the selection of the Viking landing site.



Figure 26. The radar data used to help select candidate landing sites for Viking often were expressed in degrees of rms slope.

Figure 26. The radar data used to help select candidate landing sites for Viking often were expressed in degrees of rms slope. This illustration depicts the abstract nature of that radar data. Above, (e) is the rms slope derived from the roughness data obtained near latitude -16° and shown in (b). (a) through (c) were obtained by fitting the Hagfors scattering law to measured angular power spectra, while (d) was surface reflectivity derived from data in (a) and (b). (Courtesy of Jet Propulsion Laboratory.)


The Landing Site


The selection of the final landing site of Viking 1 was a long, tedious, and dramatic process.38 The sites under consideration at the last minute, literally during the two weeks that Viking orbited Mars, all had been either observed by radar or photographed. Part of the problem was the lack of overlap between the radar and optical data. Areas with good radar data had poor photographic documentation, while sites with good photographic views had poor radar data. Everybody wanted a safe landing; nobody wanted to take a chance with a site confirmed by only radar or only photographs. The indecision foiled an earlier interest in landing Viking on the Fourth of July in honor of the country's bicentennial.

In order to acquire additional information on the candidate landing sites, the Viking Project Office commissioned another round of radar observations by Goldstone and Arecibo. Observing conditions were not ideal, because Earth and Mars were not in opposition. However, the Arecibo S-band (2380 MHz; 13-cm) 400-kilowatt radar had just come on line, and Tyler recommended making additional Arecibo observations with it. Earlier Mars radar studies had been conducted only when Mars-Earth distances were less than one astronomical unit. Signal strengths during the August 1975-February 1976 equatorial observations were good, but the Earth-Mars distance reached 2 astronomical units in May-July 1976.39

[161] Arecibo observed Mars between August 1975 and July 1976 over the latitudes between 12° South and 24° North. The results between 12° South and 4° North were relevant to potential alternate (i.e., backup) Viking sites. Between October 1975 and April 1976, Goldstone observed the two regions Syrtis Major and Sinus Meridiani, particularly a number of proposed Viking landing sites, including the prime site (called A1) near longitude 34° and latitude 19.5° North. As a result of the radar data, the A1 site was rejected on 26 June 1976, while other sites came under consideration.40

Simpson, Tyler, and Campbell made additional Arecibo observations for the Viking Project Office near 20° North latitude, the latitude of the landing site, particularly the Viking Chryse and Tritonis Lacus (the A2 site, first alternate to A1) landing areas. The search for a suitable site then moved toward the Northwest where a region designated A1NW was tentatively selected because of its apparent smoothness as seen from orbit. The A1NW site was finally abandoned because of its questionable radar properties. It was toward the west that the Viking site selection and certification teams moved after turning down A1NW.41


Did Radar Help?


Had radar observations and expressions of rms slope actually helped in the selection of the final Viking 1 landing site? Certainly, Tyler's reports to the landing site working group did not go totally unheeded, and radar turned down some potential but suspect landing sites. As NASA official John E. Naugle wrote in November 1976, "The choice of the actual landing site was eventually based on a combination of the S-band [Arecibo] radar data and high resolution photography obtained from the Viking 1 Orbiter."42 However, not everyone was as diplomatic as Naugle; some doubted the utility of the radar data.

Tom Young, Viking science integration manager, believed that radar data eventually played a role, although when the project selected initial landing site candidates, he admitted that, "radar played no role, because we weren't smart enough to know how to use it." On the other hand, James Martin, Viking Project Manager, remained skeptical about the utility of the radar data. Radar provided no useful information, he felt, although it was "an input and a source of information that [we] could not ignore." Frankly, he admitted, "The fact that it [radar] was so different scared me off."43

It was that difference, the general unfamiliarity with radar data, that raised a barrier to the use of radar results. "People didn't quite know what to make of us," Tyler explained. "People were willing to listen, but it was clear that they didn't like the answer!" Farther to the cynical side was the judgement of Dick Simpson: "I've always said that the radar contribution to picking landing sites on Mars probably came out with a net result of zero....If we'd never been involved, they probably would have had the same end result, but we got to play in the game and sometimes that's part of it." George Downs, who analyzed the Mars radar data at JPL, was convinced that project personnel simply looked for a site as Viking 1 orbited Mars, ignoring the radar data entirely. The attitude of many, he felt, was that the radar astronomers were getting their answers as if from a ouija board.44

[162] The radar data presented was indeed quite different; it was degrees of rms slope, rather than images universally understood. Perhaps if range-Doppler mapping of Mars had been possible, the difference would not have been so great. Still, the episode illustrated the kinds of challenges that radar astronomers would have to confront as they played an increasing role in planetary exploration and sought to share their results with scientists who lacked an understanding of radar. It was simply not enough to meet with planetary geologists and other scientists; radar astronomers had to communicate their results in a way understandable by other scientists.

The availability of the Mars radar data at JPL was the catalyst for the kind of interdisciplinary communication and collaboration that interpreting the radar results demanded. George Downs struck up an alliance with Ladislav Roth at JPL and Gerald Schubert at UCLA. Roth and Schubert saw value in the radar data; that is, the topographical information, not the surface roughness measurements. Roth, in fact, had approached Downs to collaborate in interpreting the radar topographical data, and several studies grew out of that collaboration.45


A Venus Radar Mapper?


Concurrently with Viking preparations, NASA planned a mission to Venus. Pioneer Venus marked a significant departure for radar astronomy. Don Campbell and Tor Hagfors had distinguished planetary radar astronomy from space exploration, in particular, the bistatic radar work done at Stanford University. Pioneer Venus challenged that distinction; it was no longer ground-based planetary radar astronomy, and it marked a significant entree into a new area of Big Science.

Instead of Big Science providing a large, Earth-based dish, like the Arecibo radar, spacecraft missions furnished the opportunity, but not the hardware, to do planetary radar astronomy from a point just above the target, not millions of kilometers away. Like piggybacking radar astronomy onto an Earth-based facility, placing a radar experiment and its necessary hardware on a spacecraft demanded participating in the politics of Big Science. Radar astronomy aboard Pioneer Venus remained Little Science, though, conducted by a single investigator, Gordon Pettengill, who carried out the entrepreneurial burden of placing the radar instrument on the spacecraft and who brought fellow ground-based radar astronomers into the project as analyzers of the radar data.

Pioneer Venus also facilitated the shift of planetary radar toward serving the planetary geology community. Within the working groups established by NASA space missions, planetary radar astronomers and planetary geologists worked together. Behind this shift, too, was the ability of radar astronomers to solve problems of interest to geologists. If planetary radar astronomy had focused solely on refining planetary orbital parameters, the prime users of planetary radar results would have remained astronomers. Radar techniques that described planetary surfaces, in contrast, solved problems of interest to geologists, especially those geologists at the United States Geological Survey (USGS) interested in planetary geology, or what the USGS called astrogeology. The shift to geology was an educational experience for both geologists and radar investigators, and it eventually manifested itself in the journals and professional societies attended by planetary radar astronomers and culminated in the Magellan mission to Venus.

The idea of using radar to image Venus from a probe predated the Pioneer Venus project. The official history of Pioneer Venus dates the beginning of the project to [163] October 1967, shortly after the Venera 4 and Mariner 5 spacecraft visited Venus. Three scientists, Richard M. Goody (Harvard University), Donald M. Hunten (University of Arizona; Kitt Peak National Observatory), and Nelson W. Spencer (Goddard Space Flight Center) formed a group to consider the feasibility of exploring the Cytherean atmosphere from a spacecraft. The group's formation led to a study published in January 1969 by the Goddard Space Flight Center.46

The idea of mapping Venus with a radar started much earlier. As early as 1959, NASA contracted with the University of Michigan to design a Venus radar. In 1961, NASA let out three more grants and contracts to develop radars for a future Venus mission to map the planet's surface to investigators at the University of New Mexico, MIT, and Ohio State.47 In 1961, for example, NASA funded a study under J. F. Reintjes, Director of MIT's Electronics Systems Laboratory, "to perform an investigation of radar techniques and devices suitable for the exploration of the planet Venus." NASA awarded the funds because the space agency saw radar as an attractive technique for exploring the surface of Venus and as "a logical experiment for a Venus flyby or orbiter."

Developing a radar system appropriate for space travel presented numerous problems. The equipment had to meet certain weight, space, and reliability criteria. The MIT goal was to design and build a space radar that required fewer than 100 watts and weighed no more than 50 pounds. After completion of an engineering model by Reintjes and the MIT Electronics Systems Laboratory, in October 1967, tests aboard an aircraft, the Convair CV-990 owned by NASA Ames Research Center, began.48

Throughout the 1960s, then, and well before the formation of the Goddard study group in 1967, the idea of imaging Venus with a spacecraft-borne radar was already "in the air." But before a spacecraft could carry a radar to Venus, NASA had to formulate and fund a voyage of exploration to the planet. In June 1968, a Space Science Board study on planetary exploration urged NASA to send a space probe to Venus, though without recommending inclusion of a radar experiment.49 By June 1970, the NASA program of planetary exploration still contained no significant Venus missions. The planned flyby of Venus and Mercury was essentially a Mercury mission with only a small contribution to Venus science. In contrast, NASA had a robust plan for exploring Mars and an ambitious program for investigating the outer planets.50

In June 1970, to address the lack of a serious Venus mission, the NASA Lunar and Planetary Missions Board and the Space Science Board brought together 21 scientists to study the scientific potential of a mission to Venus (Table 3). Richard Goody and Donald M. Hunten, who had helped start the Goddard study, co-chaired the meeting. Their report, known as the Purple Book because of the color of its cover, recommended that exploration of Venus should be prominent in the NASA program for the 1970s and 1980s. The group presented its recommendations to NASA management, and the Space Science Board endorsed them.51

Significantly, the Purple Book study brought together a planetary radar astronomer, Gordon Pettengill, then Director of the Arecibo Ionospheric Observatory, and a planetary geologist, Harold Masursky of the USGS. Pettengill's participation in the Purple Book...



Table 3. Purple Book Scientists.




Richard M. Goody, Chair

Harvard University

Donald M. Hunten

Kitt Peak National Observatory

Don L. Anderson

California Institute of Technology

W. Ian Axford

University of California, San Diego

Alan H. Barrett

Massachusetts Institute of Technology

Leverett Davis, Jr.

California Institute of Technology

Thomas M. Donahue

University of Pittsburgh

John C. Gille

Florida State University

Seymour Hess

Florida State University

Garry E. Hunt

Atlas Computer Laboratory

Robert G. Knollenberg

University of Chicago

John S. Lewis

Massachusetts Institute of Technology

Michael B. McElroy

Kitt Peak National Observatory

Gordon H. Pettengill

Arecibo Ionospheric Observatory

Robert A. Phinney

Princeton University

S. Keith Runcorn

University of Newcastle

Verner E. Suomi

University of Wisconsin

Patrick Thaddeus

Columbia University

G. Leonard Tyler

Stanford University

James A. Weiman

University of Wisconsin

George W. Wetherill

University of California, Los Angeles


....study marked his initial involvement in Pioneer Venus.52 By then, Pettengill, the future Professor of Planetary Physics in the MIT Earth and Planetary Sciences Department, had acquired stature in his field, having been one of the radar astronomy pioneers at Lincoln Laboratory, but also as Associate Director, then as Director, of the prestigious Arecibo Observatory.

Masursky had joined the USGS after graduating from Yale in 1947. After a number of years as a general geologist, Masursky joined the USGS's Branch of Astrogeologic Studies. In 1967, he became chief of the astrogeology branch, then starting in 1971 and until his death, chief scientist of that branch. Masursky was a science investigator on almost every NASA flight project to the Moon and the planets, including the Ranger, Lunar Orbiter, Surveyor, Apollo, Mariner 9, and Viking missions.53

The Purple Book meeting thus was a first step in planetary radar's shift toward geology, providing an initial setting for planetary radar and geology to interact and to develop a common approach for the study of Venus's surface, within the broader context of NASA-sponsored research of the planet's atmosphere. As the Purple Book itself noted, the space missions of the 1960s had given rise to new fields of study: "Very rapidly studies of planetary meteorology, planetary aeronomy, planetology, and planetary biology emerged which involved, in the main, research workers from the parallel terrestrial disciplines. Earth and planetary studies suddenly merged and simultaneously diverged from astronomy. In some major universities, departmental and research center organization was changed to meet this development."54

Images sent back from space had encouraged geologists, like Hal Masursky, to become interested in planetary surfaces and in the processes that shaped them. However, ground-based radar images of Venus had yet to find their audience among planetary geologists.55

[165] As far as ground-based planetary radar was concerned, the Purple Book applauded its success. "Virtually all our present knowledge of the radius, rotation, and surface of Venus has been obtained using ground-based radars," the Purple Book proclaimed. With resolutions ranging from 100 to 500 km, radar had revealed features, and even the lack of topographic relief, in the equatorial region of Venus. Including a radar system on a Venus probe would yield "maps similar in appearance and usefulness to photographic maps of the same region."

Ironically, the Purple Book cautioned against imaging Venus with a spacecraft radar. It pointed out that such radar images would be "directly competitive with ground-based observations and would provide similar data." That point resurfaced later during planning for Magellan. Although a spaceborne radar could cover more of the planet's surface, the Purple Book concluded that "it is not yet clear whether the high cost of the additional information could be justified." Not only would a spacecraft radar require "great weight and complexity" in order to compete with the resolution already achieved by ground-based radars, but the "rapidly improving capabilities of radar observatories on the earth to image Venus" made radar mapping of the planet from orbit "less important at the present time." The report reflected the anticipated benefits of the planned upgrade of the Arecibo telescope.

While technological and cost constraints militated against an orbiting radar, a viable alternative, according to the Purple Book, was the Stanford bistatic radar method, specifically that mode in which a radar on Earth transmitted and a receiver on the spacecraft collected echoes. The Purple Book concluded that "bistatic-radar experiments, in conjunction with ground-based observations, can provide a significant insight into the details of the surface structure and electromagnetic properties of Venus."

The recommendation to conduct a bistatic experiment was not surprising; Len Tyler of the SCRA was one of the 21 Purple Book scientists. Although Tyler planned to do some bistatic observations with Pioneer Venus, those plans fell by the wayside. Later, as Pettengill was writing a proposal for Pioneer Venus and was looking for scientists to join him, he invited Tyler. Tyler turned down the invitation because of his heavy commitment to the Voyager project.

In addition to the bistatic experiment, the Purple Book recommended using a radar altimeter to measure surface relief. Radar altimeter readings would complement the equatorial topographic information available from ground-based radar observations, and a simple, low-power orbiting radar could measure vertical relief over those portions of the planet not covered by ground-based radars.56 Using the altimeter to gather relief measurements was the cheapest and technologically least complicated alternative. In the end, a modified version of this approach was to fly on Pioneer Venus.

After Venera 7 succeeded in transmitting data from the surface of Venus for 23 minutes on 15 December 1970, a special panel reviewed the Purple Book conclusions. Their recommendation, to make no changes in the Purple Book, opened the door for NASA to issue an Announcement of Opportunity in July 1971 for scientists to participate in defining the Venus program.57

NASA established the Pioneer Venus Science Steering Group in January 1972, in order to enlist widespread participation of the scientific community in the early selection of the science requirements for the Pioneer Venus project. The Science Steering Group met with Pioneer Venus project personnel between February and June 1972. The Group developed in great detail the scientific rationale and objectives for several voyages to Venus and outlined candidate payloads.58

[166] The search for mission objectives stirred radar astronomer Gordon Pettengill to propose a radar experiment for the mission. Pettengill recalled: "I remember doing a calculation literally on the back of an envelope. I realized that if we could get even a tiny little antenna into a reasonable orbit around Venus, we could do an awful lot in terms of measuring the altitude and the reflecting properties of the surface....By going around Venus in a polar, rather than an equatorial orbit, we could get a totally new view of Venus. We could detail the whole surface, instead of just the equatorial band that we observed at Arecibo."

Pettengill then began "beating the drums" to include a radar experiment in the Venus program. "The Science Working Group studied the concept. I didn't think I was going to survive that," he recalled. "Pioneer Venus was strictly an atmospheric mission. A radar experiment to study the surface stood out like a sore thumb." Nonetheless, NASA awarded Pettengill funds to conduct a feasibility study of a radar to image Venus.59

Above all else, the prime mission of Pioneer Venus was to study the planet's atmosphere. An article published in 1994 in Scientific American60 evaluated the scientific achievements of Pioneer Venus and emphasized its contributions to atmospheric science, but failed to mention the radar experiment. Peter Ford, who collaborated with Pettengill on the Pioneer Venus radar experiment, pointed out that the Scientific American article's emphasis on the atmospheric science balanced the record; during the first three years of the Pioneer Venus mission, most publicity had focused on the radar imaging.61

Next, the Science Steering Group published its comprehensive report, called the Orange Book. Among the 24 areas of research advocated, only one was related to the planet's surface. As the project evolved, Pioneer Venus matured into a single-opportunity mission with a multiprobe and an orbiter. In September 1972, NASA disbanded the Science Steering Group and issued an Announcement of Opportunity for scientists to participate in the multiprobe mission. Not until August 1973 did NASA issue an Announcement of Opportunity for the orbiter. Over the ensuing months, the NASA Instrument Review Committee evaluated proposals for orbiter scientific payloads, including Pettengill's radar experiment, then presented its recommendations to NASA Headquarters in May 1974. When NASA selected the final orbiter payloads on 4 June 1974, the radar experiment was among them.

The radar was only one of 12 scientific instruments on the orbiter. In contrast to the spaceborne radar initially developed at MIT, which was to consume no more than 100 watts and weigh less than 50 pounds, the Pioneer Venus radar required only 18 watts and weighed 9.7 kilograms (21.3 pounds). "You could literally put the thing under one arm and carry it," as Pettengill characterized it. Compared to other instruments on Pioneer Venus, though, 9.7 kilograms was an appreciable load; it accounted for 22 percent of the total weight (45 kilograms) of all 12 orbiter scientific instruments.62

Although the radar experiment, in Pettengill's words, "stood out like a sore thumb," NASA Headquarters wanted to see the surface features of Venus through its white, sulfuric-acid clouds. The information was a vital part of planning for a future mission to Venus to map the planet's surface, known eventually as Magellan. The only reason the radar experiment stayed on Pioneer Venus, according to Pettengill, was that Advanced Programs at NASA Headquarters wanted it, even though its inclusion made life "a little uncomfortable for the other experiments."63

[167] The key individual in NASA's Office of Advanced Programs who supported Pettengill and the radar experiment was Daniel H. Herman. Before joining NASA in 1970 as head of Advanced Programs in the Office of Lunar and Planetary Programs, Herman had worked at Northrup on the development of surveillance synthetic aperture radar (SAR) mappers for the Navy, specifically investigating the feasibility of transmitting reconnaissance data in real time. At NASA, his job was to develop new missions and to "sell" them through the NASA hierarchy and ultimately to the President and Congress. Danny Herman's job, then, was to sell the Pioneer Venus mission. In Pettengill's words, Herman was "an eminence grise" and "a supersalesman." As early as 1972, Danny Herman also began to put together and push the Magellan project.64

Unlike Magellan, Pioneer Venus strictly speaking did not have a synthetic aperture radar; instead, the radar altimeter had a mapping mode. The most valuable data returned from the Pioneer Venus radar experiment would be the extensive topographical information acquired by the altimeter. The mapping mode did generate crude, low resolution images of portions of the planet's surface.

Far more impressive were the images generated by synthetic aperture radars (SARs) mounted on aircraft and regularly utilized by geologists to study the geology and topography of Earth. The use of SARs in Earth geology was but one part of a long and complex history that stretched from the interpretation of aerial photographs to the emergence of remote sensing, an all-encompassing term which has came to involve the interpretation of infrared, ultraviolet, microwave, gamma ray, and x-ray images, as well as optical photographs.


Radar Geology


Radar geology, as the study of geologic surface features from radar maps has come to be called, had its roots in the military surveillance radar research of the 1950s. It began to find a home in NASA during the 1960s and found a common bond with planetary radar astronomy in the 1970s, thanks largely to Pioneer Venus and Viking. The trickle of astrogeologists converted to planetary radar images by Pioneer Venus and Viking swelled through purposeful steps taken in the planning of Magellan to bring together planetary geology and planetary radar investigators.

By World War I, aerial photography had become a key tool in gathering military intelligence. The scientific applications of photointerpretation grew after the war, particularly during the 1930s. Government agencies, such as the Agricultural Adjustment Administration, the Forestry Service, and the Tennessee Valley Authority, began to use aerial photographs, and the USGS entered the field of photogrammetry, the making of maps from photographs, with a series of geologic and topographic maps constructed from aerial photographs.65

After World War II, the military sponsored research on two types of Side-Looking Airborne Radar (SLAR) used in remote sensing and especially for surveillance. One type, known as real-aperture or incoherent radar, relied on transmission of a narrow beam to provide fine image resolutions in the direction parallel to the flight of the aircraft. The other type, known as synthetic aperture radar (SAR), relied on coherent data processing to synthesize a very large effective aperture in the direction of motion and, thereby, to provide a very narrow corresponding antenna beam. Continuously operating SARs achieve a surface resolution that is independent of wavelength and approximately equal to their along-orbit physical antenna dimension. Normally, in real-aperture radars, [168] resolution is better the shorter the wavelength. In order to achieve high resolution, SARs replace the need for a large aperture with a large amount of data processing.66

The military branches developed SARs in the 1940s and 1950s under highly classified conditions in corporate and university laboratories, such as those at the Goodyear Aircraft Corporation, the Philco Corporation, the University of Illinois Control Systems Laboratory, and the University of Michigan Willow Run Research Center. By the late 1950s, a number of experimental SAR systems emerged, such as the one built by Texas Instruments for the Army. In 1961, under Air Force contract, Goodyear built the first operational SAR system; it had a resolution of about 15 meters. Throughout the 1960s, Goodyear and other firms began to commercialize SAR applications.67

A series of symposia underwritten by the Office of Naval Research (ONR) and held at the University of Michigan, where a great deal of SLAR work took place under contract with the ONR, greatly stimulated and advanced radar geology.68 The University of Michigan symposia series grew out of a study initially recommended by a subcommittee of the National Academy of Sciences (which soon formed the Committee on Remote Sensing of Environment) and the Geography Branch of the ONR. A group from the ONR and the National Academy of Sciences met in January 1961 to discuss the need for more advanced and efficient data acquisition techniques in the Earth sciences. Although University of Michigan faculty dominated the first symposium, held in February 1962, subsequent symposia participants reflected the spreading commercial importance of SAR systems in studying the Earth. By the third symposium, held in October 1964, the emphasis had shifted to remote sensing from weather and other satellites.69

During the third University of Michigan symposium, held in February 1965, R. F. Schmidt of the Avco Corporation, Cincinnati, presented a theoretical study on the feasibility of imaging Venus's surface with a radar. Schmidt failed, however, to address such practical questions as weight and power requirements.70 Nonetheless, it was clear that those interested in remote sensing, and in radar imaging in particular, were open to the idea of imaging Venus from a spaceborne radar.

Meanwhile, commercial applications of SARs to geology and topography expanded. The successful radar mapping of Panama in 1967-1968 by Westinghouse in Project RAMP, considered to be one of the major achievements in radar geology, further stimulated commercial radar mapping. In late 1971, Westinghouse surveyed the entire country of Nicaragua, and that same year the Aero Service/Goodyear RADAM Project (RADar of the AMazon), initially intended to cover only 1.5 million square kilometers, eventually covered the entire country of Brazil, over 8.5 million square kilometers. RADAM was considered the most impressive radar mapping program ever conducted.71

[169] Parallel with the development of SAR mapping of Earth was the rise of astrogeology within the USGS in the late 1950s in response to a shortage of funds and a surplus of geologists within the Survey. Following the discovery of an abundant supply of uranium ore in New Mexico, the USGS uranium project closed down in 1958. Eugene Shoemaker, a geologist who moved to the USGS Pacific Coast Regional Center at Menlo Park, California, following the closure of the uranium project, suggested lunar geologic mapping as one way to help alleviate the money and personnel problems.

Shoemaker, who did his dissertation on Meteor Crater, sold lunar geologic mapping to NASA, which in contrast to the USGS had funding but too few geologists. The result was the creation of the Astrogeologic Studies Group, USGS, Menlo Park, on 25 August 1960. Later, Shoemaker led a group of astrogeologists to a new location in Flagstaff, Arizona.72 In 1963, geologist Peter C. Badgley came to NASA from the Colorado School of Mines. Badgley was interested in techniques for observing Earth from space, particularly to support the Apollo program. He let out contracts to firms, such as Westinghouse, and universities, especially the University of Michigan, to carry out radar geologic studies from aircraft. Moreover, Badgley continued to shift NASA money to the USGS to fund lunar and planetary geology. 73 Thus, the evolution of the NASA space program and the USGS astrogeology branch marched forward in tandem.

During the Apollo program, certain USGS astrogeologists began collaborating with radar astronomers Stan Zisk and Tommy Thompson. Among them were Henry John Moore II, Shoemaker's former field assistant and part of the Menlo Park Astrogeologic Studies Group, and Gerald G. Schaber, UCLA and USGS Flagstaff.74 These early lunar efforts involved radar mapping and topographical data collected from ground-based radars, not abstract data on rms slope and dielectric constant. Schaber collaborated with Tyler on interpreting lunar bistatic radar results, which were expressed in abstract mathematical terms. Schaber admitted, "I never really did much with the interpretation of bistatic radar, because it is kind of a theoretical interpretation I don't really understand too much."75

The launch of SEASAT in the summer of 1978 began the era of satellite radar imagery. SEASAT demonstrated the feasibility of radar observations of Earth on a global basis, and initial examination of the SEASAT radar data indicated that one could fruitfully apply the data to a variety of problems in geology, agriculture, hydrology, and oceanography, as well as to planetary exploration.76

In order to assess the application of radar imaging to terrestrial geologic problems and to make recommendations to NASA, JPL sponsored the Radar Geology Workshop in Snowmass, Colorado, 16-20 July 1979, with funding from NASA. Among those on the organizing committee were Harold Masursky, USGS, and R. Stephen Saunders, JPL, who later played a role on Magellan. The workshop focused on radar observations of Earth, not the planets.77

Thus, by the launch of SEASAT in 1978, the year also of Pioneer Venus's launch, a good number of geologists were familiar with and could interpret radar images of Earth made from aircraft. But those geologists were more interested in terrestrial than extraterrestrial geology. On the other hand, through the pioneering efforts of Gene Shoemaker, the USGS Astrogeologic Studies Group already had embraced lunar radar geology. The...



Figure 27. SEASAT image of Death Valley, Earth.

Figure 27. SEASAT image of Death Valley, Earth. The launch of SEASAT in 1978 began the era of satellite radar imagery. The resolution of images made by military surveillance satellites was much finer, however. Utilization of SEASAT technology was a basic strategy adopted by JPL in the planning of VOIR (later Magellan). (Courtesy of Jet Propulsion Laboratory, photo no. P-30224.)


....potential for planetary geologists and planetary radar astronomers to work together already had been realized in the Apollo program through the work of Stan Zisk and Tommy Thompson. The NASA Pioneer Venus working committees brought together additional radar astronomers and geologists.

Once NASA decided the Pioneer Venus payloads and science experiments in June 1974, the space agency created the Orbiter Mission Operations Planning Committee. Among its members were USGS astrogeologist Hal Masursky and radar astronomer Gordon Pettengill. They also worked together closely in the Surface-Interior Working Group, one of the six mission Working Groups responsible for developing key scientific questions. Hal Masursky chaired that Working Group (Table 4).78



Table 4. Pioneer Venus Surface/Interior Working Group.




Harold Masursky

US Geological Survey

C. T. Russell

University of California, Los Angeles

Gordon H. Pettengill

Massachusetts Institute of Technology

William M. Kaula

University of California, Los Angeles

George E. McGill

Massachusetts Institute of Technology

Roger J. Phillips

Lunar and Planetary Institute

Irwin I. Shapiro

Massachusetts Institute of Technology


Pioneer Venus


Without the radar experiment, Pioneer Venus would not have brought together planetary geologists and radar astronomers. Attending meetings of the Working Groups, as well as all mission meetings, was vital to the survival of the radar on a project whose prime objectives were atmospheric. As Pettengill explained: "It was a very demanding project that had to be watched closely. I had to make sure that we did not lose radar capability. We were fighting with 11 other Principal Investigators on Pioneer Venus. It was very important that I never miss a meeting. If I missed one meeting, those guys might come to some decision that would compromise the experiment."

The Pioneer Venus atmospheric experiments competed with the radar experiment for spacecraft parameters. The atmospheric scientists wanted a different set of orbits and a different allocation of down link data bits. "It was a jungle out there!" Pettengill recalled. "You had to have a certain number of bits, or you could not do your work. If you turned your back, literally if you missed one meeting, they could make a decision to allocate 20 percent of that particular format to some experiment instead of only 10 percent. Then you have lost that 10 percent. In 1975, especially, all of this was coming together. I couldn't miss a meeting. It really was taking up my time."79

The data handling system on the orbiter integrated all analog and digital telemetry data into formats for transmission back to Earth. Telemetry storage, playback, and real-time rates varied. The orbiter had a total of 14 telemetry formats; some were used during periapsis, others during apoapsis. The radar was a heavy user of two formats designed for use at periapsis, and in fact it used more of those two formats than any other experiment.

NASA procured scientific instruments for Pioneer Venus in a variety of ways. Normally, the principal investigator was responsible for an instrument's design and construction. Either his own laboratory or a subcontractor built the instrument. NASA used a different procurement method for the Pioneer Venus radar. The project office at Ames Research Center built it for a radar team headed by Pettengill. Carl Keller, an Ames Research Center engineer, had overall decision-making responsibility, and the instrument prime contractor was the Hughes Aircraft Company Space and Communications Group, El Segundo, California, as a result of an open bid procurement. Pettengill characterized Carl Keller as an engineer from "the old school, a seat-of-the-pants, no nonsense teutonic. He would look at all the details. He was the right guy for the job. I enjoyed working with him. Not everybody did."80

[172] Both MIT's Center for Space Research, with which Pettengill was associated, and JPL competed for the Pioneer Venus radar contract. The rivalry between MIT and JPL was tense, "a real shootout" in Pettengill's words. At JPL, Walter Brown had been working on a Venus orbiter radar since the 1960s. His approach, however, differed considerably from that of MIT.

Walter Brown's radar proposal involved placing a 100-MHz (3-meter) transmitter on the Pioneer Venus orbiter, while Pettengill and the MIT Center for Space Research proposed a 1,757-MHz (17-cm) system. The MIT antenna was directive, so that when the spacecraft rotated, it took data for only a fraction of each 12-second rotation of the spacecraft.81

As Pettengill reflected: "Meanwhile, JPL thought they had the inside track. They were a NASA center, after all, and this was a NASA project. If I have a fault to lay on JPL, it is that they think that there is no place else in the world that does things as well as they do. They think they deserve the first cut of everything, because they are so much better than everybody else. They don't take kindly to new ideas that are not in-house; not invented here is very much a JPL hallmark. Irwin Shapiro has fought this on the Planetary Ephemeris Program. We fought it on radar work, and Stanford has fought it. It has been difficult over time. JPL is so institutionalized into thinking that no one else can do anything but them. It has been an uphill battle over the years. It has put grey hairs on Von Eshleman's head; it certainly put a few on mine."82

JPL lost the radar battle. Their design would have bathed the whole spacecraft, even the solar panels, in radiation from the radar. The antenna extended all around the spacecraft, so that as the spacecraft rotated, the radar always was transmitting. To Walter Brown, that was the advantage, but it made the electronics engineers nervous.

In the end, neither MIT nor JPL built the Pioneer Venus radar, but it is typical of the kinds of fights for hardware contracts that mark NASA space missions. The winner of the contract was Hughes. Hughes devised a method by which the radar altimeter could image the planet's surface at low resolution with a small, 38-cm-diameter antenna. The electronics of the MIT design were clumsy, Pettengill admitted, whereas the Hughes proposal was "very clever and efficient."

"If we had done the experiment," he mused, "it probably would not have stayed in. I have to hand Hughes some credit for that. They really had a flash of insight into a clever way of instrumenting it....They had a good team, and so did we. The main reason Hughes won was that they were willing to take a loss." For Hughes, taking a loss on the Pioneer Venus radar contract was a gambit to gain leverage on the Magellan radar contract, which they ultimately won. "At the time," Pettengill recalled, awarding Hughes the radar mapper contract "hurt a bit. I was hoping to get the hardware here at the Center for Space Research."

In August 1974, Congress approved Pioneer Venus as a new start for fiscal 1975, and in November 1974, NASA made the final contract award to Hughes Aircraft Company. By 1975, only three years away from launch, Pettengill recalled, "it all came together. With the Hughes contract, we started cutting metal."83 On 20 May 1978, the orbiter left Cape Kennedy, followed atop a second Atlas-Centaur rocket by the multiprobe on 8 August 1978. Both reached Venus in early December 1978.84

The radar was a complicated instrument capable of operating in one of two modes, altimeter or mapper. It was a 1,757-MHz (17-cm) radar with a peak output of 20 watts and utilized relatively long pulses to improve the signal-to-noise ratio. Such a radar could not [173] have performed planetary radar astronomy experiments from Earth, but reducing the distance to the target made all the difference.

Shortly after its encounter with Venus, the orbiter began making altimeter measurements of surface relief. The altimeter measured the distance from the orbiter to the planet's surface. In order to ascertain the height and depth of surface features, that distance was subtracted from the spacecraft's orbital radius, that is, the distance between the spacecraft and the planet's center of mass. The Deep Space Network, while maintaining two-way communications with the spacecraft, generated radiometric data from which JPL accurately calculated its orbit, and the MIT group then used both the orbital and radar data to determine the radius of the planet at discrete positions on the surface.

In the radar mapper mode, the instrument compensated for the complex motion of the spacecraft. Because the orbiter spun on its own axis about five revolutions per minute, radar observations took place only periodically, about one second out of each 12-second spin of the orbiter. The radar mapper also automatically compensated for the Doppler shift caused by the motion of the orbiter relative to the planet.

The instrument took altimeter data, whenever the orbiter was below 4,700 km, and imaging data, when the orbiter was below 550 km, subject to competition with other experiments for the limited telemetry capacity. In order to minimize telemetry requirements, the orbiter processed the radar echoes on board the spacecraft. The radar mapper achieved its best resolution, a footprint 23 km long and 7 km wide, at periapsis. The radar data also provided information on surface roughness and electrical conductivity.85

The radar mapper's first sweeps showed a region of Venus previously unexplored by ground-based radar. With the exception of a deep trench near the equator, the surface of Venus appeared relatively flat, similar to the Earth's surface and quite different from the rough, cratered surfaces of Mars, Mercury, and the Moon. Pioneer Venus continued to complete one orbit per day, when on the 14th orbit, the radar mapper began to malfunction; data was lost. Scientists and engineers failed to find a remedy. Mission control turned off the radar for about two weeks around Christmas and the New Year. When mission control turned on the radar mapper, they discovered that it worked, though not quite normally.

The problem, eventually traced to a timing malfunction that resulted from a differential "aging" rate in two interconnected semiconductor devices, appeared when the instrument operated longer than ten hours. Pettengill, the experiment team leader, and project personnel decided to operate the radar mapper intermittently. After about 10 days of intermittent operation, the instrument started to function normally on 20 January 1979 (orbit 47).86

Somehow, though, the mission had to recover the lost data. Data recovery was not possible during the first extended mission (September 1979), because the Deep Space Network was handling communications with Pioneer 11 at Saturn, so it took place during the second extended mission, April-May 1980. The 10 other scientific instruments (the infrared experiment failed after a few months and never ran again) continued to transmit data back to Earth; the radar mapper, however, was turned off as planned after Orbit 834 on 19 March 1981.87

[174] The processing and interpretation of Pioneer Venus altimeter and mapper data sets by the MIT group again brought together planetary geologists and radar astronomers. Peter G. Ford, Pettengill's colleague in the MIT Department of Earth and Planetary Sciences, was a central player in the MIT effort. A native of Britain, Peter Ford initially came to MIT to work in VLBI (Very Long Baseline Interferometry) radio astronomy with Irwin Shapiro. From 1977 to 1985, he worked on various aspects of the Pioneer Venus orbiting radar experiments, including their geologic interpretation, although his training was in nuclear physics. The USGS processed some of the data to create a three-dimensional effect which graphically revealed depressions and mountains. Key among the planetary geologists were Hal Masursky and Gerald Schaber of the USGS, Flagstaff, and George E. McGill, University of Massachusetts at Amherst.

The collaboration of radar astronomy and planetary geology resulted in many important discoveries about the surface of Venus, although preliminary analysis showed that much more could be learned about the planet's geological history. The altimeter data was used to create a number of maps, including a topographic contour map, a shaded relief map, and a map showing relative degrees of surface roughness. The altimeter and radar mapper data sets were assembled and placed in position by computer; however, variations from orbit to orbit were edited by hand then smoothed out by computer.

In preparation for the mission, a preliminary map was compiled from ground-based images and used by mission operations for planning. For this map, Goldstone radar images were computer mosaicked, and images obtained at Arecibo were mosaicked from photographic copy. The scale of this map was 1:50,000,000. Once the Pioneer Venus data were in hand, the map was updated to combine the spacecraft and ground-based data.

The radar altimeter yielded a topographic map covering 93 percent of the Venus globe, with a linear surface resolution of better than 150 km. Vertical measurement accuracy exceeded 200 meters. Relief was expressed as a center-of-mass-to-surface radius. Extremes went from a low of 6,049 km to a high of 6,062 km. Despite these impressive extremes of surface height and depth, the Pioneer Venus data confirmed and greatly expanded previous Earth-based observations on the global smoothness of Venus relative to the Moon, Mars, and Earth. Only about five percent of the observed surface was elevated more than two km above the mean radius, 6,051.5 ± 0.1 km.

Radar astronomer Gordon Pettengill processed and interpreted Pioneer Venus altimeter and mapper data sets. Don Campbell at Arecibo, and Dick Goldstein and Howard C. Rumsey, Jr., at JPL, supplied ground-based radar images and digital tapes, many before publication. The Arecibo and JPL radar images were compiled into a mosaic for the Pioneer Venus Planning Chart that was used in mission operations. Their high-resolution, Earth-based radar-imaging data also was essential for the interpretation of the spacecraft images and altimetry data. Thus, ground-based radar astronomers were brought into the Pioneer Venus project, and association with the project facilitated radar astronomers' access to the Goldstone radar.

The radar brightness and elevation extremes dominated the imaging and topographic maps of the highlands province, which included Ishtar Terra, Aphrodite Terra, and Beta Regio. The two highland regions, Ishtar and Aphrodite terrae, resembled terrestrial "continents" because they were high and had areas comparable to continents on Earth. Ishtar and Aphrodite appeared to be the size of continents, roughly equivalent to Australia and Africa, respectively. Beta, a much smaller feature initially detected with ground-based radar, appeared to differ from the Ishtar and Aphrodite in roughness characteristics and possibly in age and chemical composition. Ishtar Terra was the most elevated region found on Venus. It included three topographic elements: Lakshmi Planum, a western plateau area; Maxwell Montes, the central mountainous area previously studied [175] with Earth-based radars; and a complex eastern region. The highest point found on Venus was the summit of Maxwell Montes. Standing 11.1 km above the planet's average radius (in Earth terms, above sea level), Maxwell Montes was higher than Mount Everest, which reaches 8.8 km above sea level. The lowest point found on Venus was a rift valley or trench named Diana Chasma.88


Figure 28. Pioneer Venus map of Venus, 1980, showing Alpha Regio and Maxwell Montes, along the planet's meridian, and Beta Regio at longitude 280°.

Figure 28. Pioneer Venus map of Venus, 1980, showing Alpha Regio and Maxwell Montes, along the planet's meridian, and Beta Regio at longitude 280°. Diana Chasma is at longitude 160°. Compare this map with the Venus mosaic made from Arecibo Observatory radar observations (Fig. 30). (Courtesy of Jet Propulsion Laboratory, photo no. P45744.)


The planetary radar and geology collaboration yielded a host of new topographical names. In order to systematically standardize the names of Venus surface features, as well as those discovered earlier on Mars and the Moon, on an international level, the International Astronomical Union (IAU) created the Working Group for Planetary System Nomenclature (WGPSN) during its 15th General Assembly at Sydney, 21-30 August 1973. The IAU established the WGPSN because of the recent rapid advance in knowledge of the topography and surfaces of planetary bodies, as well as the necessity of coordinating the approved systems of nomenclature among the different planets and their satellites.

[176] Unlike most other IAU working groups, the WGPSN did not report through any commission or group of commissions, but was responsible to only the IAU Executive Committee. The WGPSN was charged with formulating and coordinating all topographic nomenclature on the planetary bodies of the solar system and had certain powers of action in the interval between General Assemblies. Radar astronomer Gordon Pettengill was a member of the WGPSN. The Task Group for Venus Nomenclature, responsible for compiling the detailed material presented to the WGPSN, included Gordon Pettengill, chair, JPL radar astronomer Dick Goldstein, USGS geologist Hal Masursky, and the Soviet scientist M. Ya. Marov.

Although the first meeting of the WGPSN, held in Ottawa, 27-28 June 1974, did not concern itself with the naming of surface features on Venus, at the second meeting, held in Moscow, 14-18 July 1975, the WGPSN named three valleys on Mercury Arecibo, Goldstone, and Haystack after the radar observatories and established two themes for naming Venus features. The first theme was the "feminine mystique long associated with Venus." Hence, for example, the continent-sized features Ishtar and Aphrodite were named for the Babylonian and Greek goddesses of love, respectively.

The second theme arose from the "extensive and opaque cloud cover which surrounds the planetary sphere" which "requires the use of radio and other techniques in order to study and map the surface." Therefore, the WGPSN proposed "to assign the names of deceased radio, radar and space scientists to topographic features." One exception, Alpha, was admitted. Alpha was one of the first Cytherean features to be observed "and which has served to help define the origin of the official IAU system of longitude for the planet." During subsequent meetings of the WGPSN, held in Grenoble, 30-31 August 1976; Washington, 1-2 June 1977; Innsbruck, 2 June 1978; and Montreal, 13-15 August 1979, the WGPSN approved not only Alpha, but Beta and Maxwell as well.89 Thus, the feature names first given by ground-based radar astronomers were fixed on the map of Venus.

Pioneer Venus awakened more planetary geologists to the value of radar data, especially radar images. Pioneer Venus also was a new taste of Big Science that would lead to the Magellan mission. In turn, Magellan culminated the linking of planetary geology with radar astronomy and further blurred the distinction made earlier in the history of planetary radar astronomy between ground-based radar and space exploration.


1. Smith and Carr, pp. 130-131.

2. Kuznetsov and Lishin, p. 201.

3. C. W. Snyder, "Mariner 5 Flight past Venus," Science 158 (1967): 1665-1669; Arvydas Kliore, Gerald S. Levy, Dan L. Cain, Gunnar Fjeldbo, S. Ichtiaque Rasool, "Atmosphere and Ionosphere of Venus from the Mariner 5 S-band Radio Occultation Experiment," Science 158 (1967): 1683-1688; Gerard H. de Vaucouleurs and Donald H. Menzel, "Results of the Occultation of Regulus by Venus, July 7, 1959," Nature 188 (1960): 28-33; Ash, Shapiro, and Smith, Astronomical Journal 72 (1967): 338-350.

4. Ash, Campbell, Dyce, Ingalls, Jurgens, Pettengill, Shapiro, Martin A. Slade, and Thompson, "The Case for the Radar Radius of Venus," Science 160 (1968): 985-987; Ash, Campbell, Dyce, Ingalls, Jurgens, Pettengill, Shapiro, Slade, Smith, and Thompson, "The Case for the Radar Radius of Venus," Journal of the Atmospheric Sciences 25 (1968): 560-563; Shapiro 1/10/93.

5. William G. Melbourne, Muhleman, and D. A. O'Handley, "Radar Determination of the Radius of Venus," Science 160 (1968): 987-989.

6. Kliore and Cain, "Mariner 5 and the Radius of Venus," Journal of Atmospheric Sciences 25 (1968): 549-554; Campbell 7/12/93. Murray, pp. 90-91, provides further anecdotal accounting of Soviet embarrassment over the incident.

7. Goldstein and Gillmore, "Radar Observations of Mars," Science 141 (1963): 1172.

8. Goldstein, "Mars: Radar Observations," Science 150 (1965): 1715-1717. His results were reported also in Goldstein, "Preliminary Mars Radar Results," Radio Science 69D (1965): 1625-1627.

9. Dyce, Pettengill, and Sánchez, "Radar Observations of Mars and Jupiter at 70 cm," The Astronomical Journal 72 (1967): 771-777; Campbell 7/12/93.

10. Carl Sagan, James B. Pollack, and Goldstein, "Radar Doppler Spectroscopy of Mars: 1. Elevation Differences between Bright and Dark Areas," The Astronomical Journal 72 (1967): 20-34. This article appeared earlier as Sagan, Pollack, and Goldstein, Radar Doppler Spectroscopy of Mars: 1. Elevation Differences between Bright and Dark Areas, Special Report 221 (Cambridge: SAO, 6 September 1966).

11. See, for example, D. G. Rea, "The Darkening Wave on Mars," Nature 210 (1964): 1014-1015; R. A. Wells, "Evidence that the Dark Areas on Mars are Elevated Mountain Ranges," Nature 207 (1965): 735-736. Rea was at the University of California at Berkeley, and Wells at University College, London.

12. Pettengill, Counselman, Rainville, and Shapiro, "Radar Measurements of Martian Topography," The Astronomical Journal 74 (1969): 461-482; Pettengill, Rogers, and Shapiro, "Martian Craters and a Scarp as Seen by Radar," Science 174 (1971): 1324.

13. Gold, "The Lunar Surface," Monthly Notices of the Royal Astronomical Society 115 (1955): 585-604; Malcolm J. Campbell, Juris Ulrichs, and Gold, "Density of the Lunar Surface," Science 159 (1968): 973; Gold and Steven Soter, "Apollo 12 Seismic Signal: Indication of a Deep Layer of Powder," Science 169 (1970): 1071-1075; Gold, "The Moon's Surface," in Wilmot N. Hess, Menzel and John A. O'Keefe, eds., The Nature of the Lunar Surface (Baltimore: Johns Hopkins University Press, 1966), pp. 107-121; Gold, "Conjectures about the Evolution of the Moon," The Moon 7 (May-June 1973): 293-306.

14. Don E. Wilhelms, To A Rocky Moon: A Geologist's History of Lunar Exploration (Tucson: The University of Arizona Press, 1993), p. 347.

15. Wilhelms, p. 299.

16. Gold 14/12/93.

17. Schaber 27/6/94; Thompson 29/11/94.

18. Corliss, The Viking Mission to Mars, NASA SP-334 (Washington: NASA, 1974), pp. 6-8; Thomas A. Mutch, Raymond E. Arvidson, James W. Head, III, Kenneth L. Jones, R. Stephen Saunders, The Geology of Mars (Princeton: Princeton University Press, 1976).

19. Martin Marietta Aerospace, The Viking Mission to Mars (Denver: Martin Marietta, 1975), pp. III-21 to III-23; Edward Clinton Ezell and Linda Neuman Ezell, On Mars: Exploration of the Red Planet, 1958-1978, NASA SP-4212 (Washington: NASA, 1984), p. 298.

20. Campbell 8/12/93.

21. Tyler 10/5/94; Ezell and Ezell, pp. 309 & 320-321; "VOIR, Proposal to the NASA Management Section, 2/79," Box 13, JPLMM.

22. SCRA, Research at the Stanford Center for Radar Astronomy, semi-annual status report no. 2 for the period 1 July - 31 December 1963 (Stanford: RLSEL, February 1964), pp. 3-4; Ibid., no. 4 for the period 1 July - 31 December 1964 (Stanford: RLSEL, January 1965), pp. 2-3; Ibid., no. 5 for the period 1 January - 30 June 1965 (Stanford: RLSEL, July 1965), pp. 5-6; Ibid., no. 6 for the period 1 July - 31 December 1965 (Stanford: RLSEL, January 1966), p. 4; Ibid., no. 7 for the period 1 January - 31 June 1966 (Stanford: RLSEL, August 1966), p. 5; Ibid., no. 9 for the period 1 January - 30 June 1967 (Stanford: RLSEL, 9 July 1967), pp. 6-8; John E. Ohlson, A Radar Investigation of the Solar Corona, SU-SEL-67-071, Scientific Report 21 (Stanford: RLSEL, August 1967).

23. Simpson 10/5/94.

24. Richard A. Simpson and G. Leonard Tyler, "Viking Bistatic Radar Experiment: Summary of First-Order Results Emphasizing North Polar Data," Icarus 46 (1981): 361-389; Simpson and Tyler, "Radar Measurement of Heterogeneous Small-Scale surface Texture on Mars: Chryse," Journal of Geophysical research 85 (1980): 6610-6614; Simpson 10/5/94.

25. Fjeldbo, "Bistatic-Radar Methods for Studying Planetary Ionospheres and Surfaces," Ph.D. diss., Stanford University, 1964, especially pp. 64-82. Later published as Fjeldbo, Bistatic-Radar Methods for Studying Planetary Ionospheres and Surfaces, SR 2 (Stanford: RLSEL, 1964).

26. Tyler 10/5/94; Tyler, The Bistatic Continuous-Wave Radar Method for the Study of Planetary Surfaces, SU-SEL-65-096, Scientific Report 13 (Stanford: RLSEL, October 1965), which later appeared as Tyler, "The Bistatic, Continuous-Wave Radar Method for the Study of Planetary Surfaces," Journal of Geophysical Research 71 (1966): 1559-1567; Tyler, Bistatic-Radar Imaging and Measurement Techniques for the Study of Planetary Surfaces, SU-SEL-67-042, Scientific Report 19 (Stanford: RLSEL, May 1967); Tyler and Simpson, Bistatic-Radar Studies of the Moon with Explorer 35: Final Report Part 2, SR 3610-2, SU-SEL-70-068 (Stanford: RLSEL, October 1970); Tyler and Simpson, "Bistatic Radar Measurements of Topographic Variations in Lunar Surface Slopes with Explorer 35," Radio Science 5 (1970): 263-271; SCRA, Proposal to the National Aeronautics and Space Administration for Bistatic Radar Astronomy Studies of the Surface and Ionosphere of the Moon based upon Transmission from the Earth and Reception in a Surveyor Orbiter, Proposal RL 21-62 (Stanford: RLSEL, 7 September 1962), Eshleman materials.

27. Tyler 10/5/94; Simpson 10/5/94; Simpson and Tyler, "Radar Scattering Laws for the Lunar Surface," IEEE Transactions on Antennas and Propagation AP-30 (1982): 438-449; Simpson, "Lunar Radar Echoes: An Interpretation Emphasizing Characteristics of the Leading Edge," Ph.D. diss., Stanford University, 1973.

28. Tyler 10/5/94; Rogers, Ash, Counselman, Shapiro, and Pettengill, "Radar Measurements of Surface Topography and Roughness of Mars," Radio Science 5 (1970): 465-473; Goldstein, Melbourne, Morris, George S. Downs, and O'Handley, "Preliminary Radar Results of Mars," Radio Science 5 (1970): 475-478.

29. Goldstein 14/9/93; Downs, Goldstein, R. Green, Morris, "Mars Radar Observations, A Preliminary Report," Science 174 (1971): 1324-1327; Downs, Goldstein, R. Green, Morris, and Reichley, "Martian Topography and Surface Properties as Seen by Radar: The 1971 Opposition," Icarus 18 (1973): 8-21; Pettengill, Shapiro, and Rogers, "Topography and Radar Scattering Properties of Mars," Icarus 18 (1973): 22-28; Pettengill, Rogers, and Shapiro, "Martian Craters and a Scarp as Seen by Radar," Science 174 (1971): 1321-1324.

30. Pettengill, Shapiro, and Rogers, "Topography and Radar Scattering Properties of Mars," Icarus 18 (1973): 22-28; Pettengill, Rogers, and Shapiro, "Martian Craters and a Scarp," pp. 1321-1324.

31. Simpson, Tyler, and Belinda J. Lipa, Analysis of Radar Data from Mars, SR 3276-1, SU-SEL-74-047 (Stanford: SCRA, October 1974).

32. Ingalls 5/5/94; Simpson 10/5/94.

33. Memorandum, Sebring to Distribution, 9 December 1970, 44/2/AC 135; "Applications of High Power Radar to Studies of the Planets, NASA, 7/1/69-6/30/70," 67/2/AC 135; "Radar Studies of the Planets, NASA, 7/1/72-6/30/73," 68/2/AC 135, MITA; NEROC, Final Progress Report Radar Studies of the Planets, 29 August 1974, pp. 1-2; NEROC, Semiannual Report of the Haystack Observatory, 15 July 1972, p. ii; Simpson, Tyler, and Lipa, "Mars Surface Properties Observed by Earth-Based Radar at 70-, 12.5-, and 3.8-cm Wavelengths," Icarus 32 (1977): 148. For the radar results themselves, see Pettengill, John F. Chandler, Campbell, Dyce, and D. M. Wallace, "Martian Surface Properties from Recent Radar Observations," Bulletin of the American Astronomical Society 6 (1974): 372; Downs, Goldstein, R. Green, Morris, and Reichley, "Martian Topography and Surface Properties as Seen by Radar: The 1973 Opposition," Icarus 18 (1973): 8-21; Downs, Reichley, and R. Green, "Radar Measurements of Martian Topography and Surface Properties: The 1971 and 1973 Oppositions," Icarus 26 (1975): 273-312.

34. Ezell and Ezell, p. 298.

35. Simpson, Tyler, and Lipa, "Mars Surface Properties Observed by Earth-Based Radar at 70-, 12.5-, and 3.8-cm Wavelengths," Icarus 32 (1977): 156.

36. Ezell and Ezell, p. 322.

37. Simpson 10/5/94; Schaber 27/6/94; Shoemaker 30/6/94; Soderblom 26/6/94; Gold 14/12/93.

38. For a full discussion, see Ezell and Ezell, pp. 317-346, as well as Downs 4/10/94.

39. Simpson, Tyler, and Lipa, Analysis of Radar Data from Mars; Simpson, Tyler, and Campbell, "Arecibo Radar Observations of Mars Surface Characteristics in the Northern Hemisphere," Icarus 36 (1978): 156-157.

40. Simpson, Tyler, and Campbell, "Arecibo Radar Observations of Martian Surface Characteristics Near the Equator," Icarus 33 (1978): 102-115; Downs, R. Green, and Reichley, "Radar Studies of the Martian Surface at Centimeter Wavelengths: The 1975 Opposition," Icarus 33 (1978): 441-453.

41. Simpson, Tyler, and Campbell, "Mars Surface Characteristics in the Northern Hemisphere," pp. 153-173.

42. John E. Naugle to H. Guyford Stever, 8 November 1976, NHOB.

43. Ezell and Ezell, p. 357.

44. Tyler 10/5/94; Simpson 10/5/94; Downs 4/10/94.

45. Downs 4/10/94; Ladislav E. Roth, Downs, Saunders, and Gerald Schubert, "Radar Altimetry of South Tharsis, Mars," Icarus 42 (1980): 287-316; Roth, Saunders, Downs, and Schubert, "Radar Altimetry of Large Martian Craters," Icarus 79 (1989): 289-310.

46. Richard O. Fimmel, Lawrence Colin, and Eric Burgess, Pioneer Venus, NASA SP-461 (Washington: NASA, 1983), pp. 14-15; Colin, "The Pioneer Venus Program," Journal of Geophysical Research 85 (1980): 7575.

47. Tatarewicz, pp. 150-151.

48. Memorandum, Oran W. Nicks, 10 March 1966, and Memorandum, Brunk, 29 November 1966, NHOB; J. F. Reintjes and J. R. Sandison, Venus Radar Systems Investigations Final Report (Cambridge: MIT, Electronic Systems Laboratory, Department of Electrical Engineering, March 1970), Pettengill materials.

49. Space Science Board, Planetary Exploration, 1968-1975 (Washington: National Academy of Sciences, 1968).

50. Space Science Board, Venus: Strategy for Exploration (Washington: National Academy of Sciences, June 1970), p. 3.

51. C. H. Townes, Preface, Space Science Board, Venus: Strategy for Exploration, n.p.

52. Pettengill 28/9/93.

53. V-Gram no. 12 (July 1987): 15; "Harold Masursky," in R. R. Bowker, comp., American Men and Women of Science, 18th edition (New Providence, NJ: R. R. Bowker, 1992), vol. 5, p. 275.

54. Space Science Board, Venus: Strategy for Exploration, p. 4.

55. Campbell 9/12/93.

56. Tyler 10/5/94; Simpson 10/5/94; Venus: Strategy for Exploration, pp. 58-62.

57. Fimmel, Colin, and Eric, pp. 17-18.

58. Fimmel, Colin, and Burgess, p. 18.

59. Pettengill 28/9/93.

60. Janet G. Luhmann, Pollack, and Colin, "The Pioneer Mission to Venus," Scientific American 270 (April 1994): 90-97.

61. Ford 3/10/94.

62. Pettengill 28/9/93; Fimmel, Colin, and Burgess, pp. 18-21, 38 & 58.

63. Pettengill 28/9/93.

64. Pettengill 28/9/93; Daniel H. Herman, telephone conversation, 20 May 1994.

65. William A. Fischer, "History of Remote Sensing," in Robert G. Reeves, Manual of Remote Sensing (Falls Church, Virginia: American Society of Photogrammetry, 1975), [2 volumes] vol. 1, pp. 27-39.

66. For a discussion of space SARs by one of its leading practitioners, see Charles Elachi, Spaceborne Radar Remote Sensing: Applications and Techniques (New York: IEEE Press, 1987).

67. Fischer, pp. 42-43; Allen M. Feder, "Radar Geology, the Formative Years," Geotimes vol. 33, no. 11 (1988): 11-14. See also John J. Kovaly, Synthetic Aperture Radar (Dedham, MA: Artech House, Inc., 1976), Chapter One. I am grateful to Louis Brown for this last reference.

68. Feder, p. 12.

69. Proceedings of the First Symposium on Remote Sensing of Environment (Ann Arbor: University of Michigan Institute of Science and Technology, March 1962). Of the 72 participants, 37 of them, or 51%, were University of Michigan faculty. Proceedings of the Second (Ann Arbor: University of Michigan Institute of Science and Technology, February 1963); Proceedings of the Third Symposium (Ann Arbor: University of Michigan Institute of Science and Technology, October 1964); Proceedings of the Fourth Symposium (Ann Arbor: University of Michigan Institute of Science and Technology, June 1966).

70. Peter C. Badgley, "The Applications of Remote Sensors in Planetary Exploration," in Proceedings of the Third Symposium, pp. 9-28; R. F. Schmidt, "Radar Mapping of Venus from an Orbiting Spacecraft," ibid., pp. 51-61.

71. Feder, p. 13; H. MacDonald, "Historical Sketch: Radar Geology," pp. 23-24 & 27-28 in Radar Geology: An Assessment Publication 80-61 (Pasadena: JPL, 1 September 1980). This was a report of the Radar Geology Workshop, held at Snowmass, Colorado, 16-20 July 1979.

72. Wilhelms, pp. 37-40, 43, 46, 48 & 77.

73. Pamela E. Mack, Viewing the Earth: The Social Construction of the Landsat Satellite System (Cambridge: The MIT Press, 1990), pp. 46-49; MacDonald, pp. 26 & 28-29.

74. Wilhelms, p. 47; Thompson 29/11/94. See, for instance, Shapiro, Stanley H. Zisk, Rogers, Slade, and Thompson, "Lunar Topography: Global Determination by Radar," Science 17 (1972): 939-948.

75. Schaber 27/6/94.

76. John P. Ford, "Seasat Orbital Radar Imagery for Geologic Mapping: Tennessee-Kentucky-Virginia," American Association of Petroleum Geologists Bulletin 64 (1980): 2064-2094; Radar Geology: An Assessment, p. 1.

77. Radar Geology: An Assessment, passim.

78. Fimmel, Colin, and Burgess, pp. 22 & 218.

79. Pettengill 28/9/93.

80. Pettengill 28/9/93; Fimmel, Colin, and Burgess, pp. 22, 41 & 43.

81. Pettengill 28/9/93; Memorandum, Brunk, 29 November 1966, NHOB.

82. Pettengill 28/9/93.

83. Pettengill 28/9/93.

84. Fimmel, Colin, and Burgess, pp. 27 & 35.

85. Fimmel, Colin, and Burgess, pp. 58-59 & 113-115; Pettengill, D. F. Horwood, and Carl H. Keller, "Pioneer Venus Orbiter Radar Mapper: Design and Operation," IEEE Transactions on Geoscience and Remote Sensing GE-18 (1980): 28-32; Pettengill, Peter G. Ford, and Stewart Nozette, "Venus: Global Surface Radar reflectivity," Science 217 (1982): 640-642.

86. Pettengill, Ford, Walter E. Brown, William M. Kaula, Carl H. Keller, Harold Masursky, and George E. McGill, "Pioneer Venus Radar Mapper Experiment," Science 203 (1979): 806-808; Colin, "The Pioneer Venus Program," Journal of Geophysical Research 85 (1980): 7588-7589; Fimmel, Colin, and Burgess, p. 107; Pettengill, Ford, Brown, Kaula, Masursky, Eric Eliason, and McGill, "Venus: Preliminary Topographic and Surface Imaging results from the Pioneer Orbiter," Science 205 (1979): 91-93.

87. Colin, pp. 7589 & 7590; Fimmel, Colin, and Burgess, p. 191.

88. Fimmel, Colin, and Burgess, p. 154; Masursky, Eliason, Ford, McGill, Pettengill, Gerald G. Schaber, and Schubert, "Pioneer Venus Radar Results," Journal of Geophysical Research 85 (1980): 8232-8260; Pettengill, Eliason, Ford, George B. Loriot, Masursky, and McGill, "Pioneer Venus Radar Results: Altimetry and Surface Properties," Journal of Geophysical Research 85 (1980): 8261-8270; V-Gram no. 10 (January 1987): 20.

89. Transactions of the International Astronomical Union 17A (1979): 113-114; "Working Group for Planetary System Nomenclature," Ibid. 16B (1977): 321-369; "Working Group for Planetary System Nomenclature," Ibid. 17B (1980): 285-304.