Magellan Bistatic Radar Experiments Experiment Description During Magellan's primary mission, radar images of Venus' surface were acquired using the on-board synthetic aperture radar system. Data collected during all three mapping cycles covered more than 99% of the surface at resolutions as fine as 75 m. During the same mapping operations, the nadir-looking altimeter obtained topography between latitudes 85N and 85S with vertical and horizontal resolutions on the order of 10 m and 10 km, respectively. Although the image and altimetry data represent a substantial foundation from which geologists and geophysicists can improve our understanding of Venus, it is important to remember that the SAR images are not photographs and that radar altimetry is not equivalent to surveying. In some respects--particularly as regards uniformity of coverage--radar techniques can be both faster and of higher quality than alternative methods. In other respects, the use of radar can lead to subtly different (and often confusing) results that are difficult to interpret. One outstanding puzzle from the radar missions to Venus is the anomalously high radar reflectivity observed near the summits of most mountains. Another is the apparent anisotropy of radar scattering from certain regions on the surface. Magellan bistatic radar observations, conducted in an entirely different experimental geometry, may shed some new light on these unusual results. Anomalous Reflectivity The reflectivity enhancements (and corresponding suppression of emissivity) were first reported during analysis of Pioneer Venus Orbiter (PVO) data. At most elevations above 6054 km, reflectivities increase sharply from nominal Venus values of 0.15 to 0.4 and more. The transition at 6054 km level is not ubiquitous, but it seems to apply for a surprisingly large fraction of the planet. There are a small number of exceptions--high regions that do not show enhanced reflectivity and low regions which do. Interpretations were originally based on the fact that "loaded" dielectrics--matrices of insulating material in which small nodules of metal are distributed--can produce the observed effects. Pyrite, for example, was an early candidate--inclusions of metallic material modified its otherwise benign properties so that it sometimes behaves as though it has a dielectric constant on the order of 50 (by comparison, most soils are less than 3 and most rocks are less than 10). Subsequent calculations and laboratory results have indicated that pyrite would not be stable in the Venus environment over geologic time scales, but other minerals with similar properties have since been proposed. An alternative explanation for the high reflectivity is based on the fact that inhomogeneities in very low loss materials can produce scattering. Within a dielectric material, voids having dimensions comparable to the radar wavelength can cause the electromagnetic wave to be redirected in such a way that much of the incident energy is scattered back into space. A variation on this model includes a symmetry argument--that photons traveling over the same path through the surface material, but in opposite directions, will return to the radar in phase. This coherence can yield a much higher echo power than in the case when the photons reach the radar with random phases. In the backscatter direction (and a narrow cone around that direction) there will be enhanced echo power so long as the scattering is accomplished by the inhomogeneities before the wave is attenuated in the surface material. Such a mechanism appears to control radar backscatter from icy surfaces on the Galilean satellites, Mars, and Mercury. Whether a similar mechanism could act on Venus--where water ice is an unlikely candidate--remains a subject of active discussion. Limitations of Monostatic Observing Geometry All Magellan radar probing of Venus during its primary mission was done in the backscatter configuration (DESC_FG1.EPS). If enhanced backscatter is contributing to the unusual echo powers in the Magellan (and PVO) data, then the measured reflectivity must be adjusted downward before the Fresnel relationships are applied to derive dielectric constant. The difficulty comes in deciding which data include a coherently enhanced component, since all measurements have been carried out using the same (backscatter) geometry. If loaded dielectrics are responsible for the enhanced reflectivity, then no correction should be applied and estimates of dielectric constant derived directly from the reflectivity will be accurate. The different implications of these two models for composition of the surface material are significant. A variation on the normal Magellan experimental geometry was used for orbits 3716-3719. Although the configuration was still backscatter, the spacecraft was rotated 90 deg about the boresight of its high gain antenna. In this way, the polarization of the antenna was changed from "horizontal" to "vertical." The scattering of horizontally and vertically polarized waves is different, but the details are not well- understood theoretically. Instead, the objective for these orbits was to observe the vertically polarized emissions from the surface so those could be compared with the horizontally polarized emissions from the same terrain on previous orbits. If the surface were a relatively smooth dielectric, then the emissivity of the surface would change in a predictable way--based on the Fresnel relationships for a dielectric interface. The results from those special observations have proven to be as difficult to interpret as the original data--the emissivity varies qualitatively in the way predicted by theory, but the quantitative results have not been "satisfactory," leading some to conclude that the simple model of a dielectric interface (loaded or not) should be ruled out. Bistatic Geometry Another way of studying the anomalous scattering is to abandon the monostatic (DESC_FG1.EPS) geometry and to investigate the scattering and emission behavior of the surface under a much wider variety of angular conditions. Spotlight Mode: Earlier in the Magellan Mission Gordon Pettengill had proposed identifying a target area on Venus that could be observed in a "spotlight" bistatic configuration. The spacecraft would continuously illuminate a fixed surface region as it flew over, while the echo signal would be captured on Earth (DESC_FG2.EPS). Scattering could be observed at a wide range of angles; if the target and the timing were chosen properly, the backscatter geometry would also be included in the same observations. Gula Mons, the summit of which shows a modest enhancement in Magellan reflectivity, was chosen for such a test on 6 October 1993. The spacecraft was programmed to point its high-gain antenna toward the summit for slightly more than 8 minutes on 6 consecutive orbits. The Magellan Project was able to schedule 70-m coverage for one of the orbits from Goldstone and for two of the later orbits from Canberra. Good data from the first Canberra orbit were processed, but no echo has been identified. Because of the large Venus-Earth distance, the echo was expected to be only a few times larger than the measurement uncertainties. Refinements to the processing may still yield a detection, and that work continues. A recording failure resulted in loss of at least half of the Goldstone data, while the second Canberra orbit was corrupted by an unknown modulation. In both cases it may be possible to recover the samples, but further work is needed to understand the details of the failures. Quasi-Specular Mode: A second alternative geometry is to conduct bistatic observations of the quasi-specular echo. In this configuration, the spacecraft high gain antenna is aimed at the point on Venus' mean spherical surface expected to give mirror-like reflection toward Earth (DESC_FG3.EPS). It is the quasi-specular echo that the Magellan altimeter measured, since the specular component dominates the nadir reflection. As the spacecraft moves in the bistatic configuration, the specular point follows on a parallel path. When the orbit is viewed edge-on from Earth, the track is identical to that of the sub-spacecraft track (though slightly ahead or behind); when the orbit is viewed face-on, the specular track mimics the orbit ellipse but is smaller (DESC_FG4.EPS). The width of the echo spectrum can be related to the rms surface slopes; the strength of the spectrum is proportional to the reflectivity, which can then be translated to dielectric constant (DESC_FG5.EPS). Three Magellan orbits on 9 November 1993 were used to obtain quasi- specular bistatic echoes; data were collected using the Madrid 70-m DSN antenna. Through the real-time monitoring systems at JPL it was possible to observe intermittent echoes in real time during these tests. The 70-m facilities are equipped to record both S- and X-Band data in both right- and left-circular polarizations; echoes were observed only at S-band. Recording problems compromised some of the data, but several complete orbits were obtained before the failures began. One of the real-time monitor spectra is shown in DESC_FG6.EPS; it's half-power width is consistent with an rms slope of 0.5 deg at (25N, 69E). With two (independent) S-Band channels, it is possible to make independent estimates of quantities such as rms slope. Maxwell Montes: One of the largest regions of enhanced backscatter is the highly elevated region known as Maxwell. A fortunate coincidence allowed scheduling of a 'spotlight' experiment at one target on Maxwell on 31 May 1994 and quasi-specular observations over the same area six days later. In fact the quasi-specular track extended from near the equator to well north of Maxwell, allowing comparison of scattering among many terrain types. The spotlight experiment was conducted using the Madrid 70-m DSN antenna; the quasi-specular observations were conducted first using the Madrid facility and later using the Goldstone antenna. Most of the Madrid data appear to be good, though no spotlight echo has been detected and only S-band quasi-specular echoes have been seen. A ground antenna pointing error severely compromised the Goldstone observations, making them far less useful than the Madrid data. By processing the left- and right-circularly polarized quasi- specular data at S-band coherently, Pettengill et al. (Science, 272, 1628- 1631, 1996) were able to infer a complex dielectric constant from the region with enhanced backscatter. In fact, the complex dielectric constant matches the behavior one would expect from a thin layer of the semiconductor tellurium under Venus conditions. Analysis: For much of the surface the limits of the high-gain antenna beam pattern (rather than the intrinsic roughness of the surface) will dominate the observed echo shape. Extracting rms slopes and dielectric constant requires modeling the scattering geometry. The emphasis above has been on the value of Magellan bistatic experiments to unravel one of the outstanding scattering questions about Venus' surface--the anomalous backscatter reflectivities at high Venus altitudes. The question of anisotropic scattering, raised in the introductory paragraphs, has not been explored here but may also be addressed using these data in conjunction with monostatic Magellan data. Figure Captions: Figure 1 (DESC_FG1.EPS): During Magellan's mapping operations, SAR data were collected over a narrow range of angles (~2 deg) at incidence angles as large as 45 deg at periapsis to about 15 deg over the poles. The nadir-viewing altimeter also collected backscatter echoes, but at angles extending only to about 10 deg from vertical. Figure 2 (DESC_FG2.EPS): In a spotlight configuration, the radar is pointed at a single location on the surface as the spacecraft flies over. No monostatic spotlight experiments were conducted with Magellan; the first bistatic spotlight experiment was conducted over Gula Mons on 6 October 1993. Figure 3 (DESC_FG3.EPS): In the quasi-specular mode, the spacecraft antenna is aimed at the point which gives mirror-like reflection toward the receiver--on Earth in this case. As the spacecraft moves, the antenna tracks the specular point and the incidence angle at the specular point usually varies. Figure 4 (DESC_FG4.EPS): Extreme viewing geometries for quasi- specular observations occur when the orbit is viewed edge-on and face- on. The edge-on condition provides the widest range of incidence angles but also brings the most dynamic signal behavior. The face-on condition allows study of only the higher incidence angles but is easier to implement from a planning and data acquisition point of view. Figure 5 (DESC_FG5.EPS): The central part of the quasi-specular echo arises from mirror-like reflection at the specular point. The echo is broadened because surface facets displaced from the specular point and slightly tilted with respect to the mean surface can also give mirror-like reflections. The amount of broadening is a measure of the rms tilts on the surface; the strength of the echo can be related to the Fresnel reflection coefficient (and, hence, the dielectric constant) of the surface material. If the spacecraft antenna beam is too narrow to illuminate all the potential contributors to the quasi-specular echo, the echo shape will be determined by the antenna beam rather than the surface rms tilt. Figure 6 (DESC_FG6.EPS): One of the strongest quasi-specular echoes from the Magellan experiments of 9 November yielded an echo peak amplitude 6-7 dB above the noise baseline. The surface is smooth enough that the antenna beamwidth is not a factor and a surface rms slope of 0.5 deg can be estimated from the 493 Hz half-power bandwidth. A surface location (25N, 69E) is based on early planning calculations.