When an electromagnetic wave is transmitted toward a surface, the amount of energy which is scattered from the surface back to the transmitter is most heavily dependent upon the ``roughness'' or ``texture'' of the surface on size scales of the order of the wavelength of the incident electromagnetic wave. Measurement of the reflected power yields information about the radar reflectivity - and therefore the composition - of the surface. Comparison of the scattering seen in the sense of circular polarization opposite to that transmitted (flat-facet reflections from a dielectric) to that seen in the same sense (caused by multiple reflections or by certain types of ices) yields additional information about surface roughness on scales comparable to the wavelength. So, for the 8.4 GHz band, surface structures from slightly sub-cm to some 10's of cm are most important, and dictate the amount of received flux. Going to longer wavelengths (such as the 2.4 GHz band) provides information on surface roughness at larger scales, approaching 1 m.
Since the early 1960's, results from planetary radar experiments have been used to help form a more complete understanding of the surfaces of the terrestrial planets. Since the installation of the 8.4 GHz band receivers on the antennas of the VLA in 1988, it has become possible to use it in combination with the Goldstone 70 m antenna to create the most powerful radar instrument in the world. With the Goldstone/VLA instrument, it is possible to image the surfaces of Venus, Mercury and Mars with a spatial resolution as good as 100 km. It is also possible to probe the surfaces of the Galilean satellites and Titan. The results of such experiments are some of the most exciting in recent years in planetary science. Perhaps the most significant - and most unanticipated - of these results was the discovery of ice deposits in the polar regions of Mercury. Other extremely important results came from probing of the ice caps of Mars and finding them mysteriously different from each other, discovery of a huge region near the equator of Mars which reflects no detectable energy (dubbed ``Stealth''; Fig. 2.3), and the discovery that Titan has no deep global ethane/methane ocean - contrary to theoretical predictions. The Goldstone/VLA instrument has also been used to image several near-earth asteroids, and an experiment on a comet is planned for the future. While these results are impressive, and similar experiments can and will continue to be performed, the VLA Development Plan will have a major impact on planetary science.
Figure 2.3: This is a simple cylindrical projection of values of the radar cross section of Mars at normal incidence in the same sense polarization. Darker shades represent higher values of cross section. These values are obtained from a fit to the backscatter function of all data points for a particular location on the surface. The region of ``Stealth'' is shown by the dark outlines. The solid contour encloses locations where the cross section at normal incidence is less than the noise value. The broken outline is an earlier estimate, from 1988 data only. White outlines denote the approximate geological boundaries of the calderas and shields of the Tharsis volcanoes.
First, the addition of the 2.4 GHz band would be of major interest. Both the Goldstone 70 m, and the Arecibo 305 m antennas, are equipped with powerful 2.4 GHz transmitters, and could be used in combination with the VLA. While the upgraded Arecibo system will have more power ( MW), the Goldstone transmitter ( KW) has full-sky coverage, which is particularly important for radar studies of comets and asteroids. Radar observations at 2.4 GHz, in combination with 8.4 GHz measurements already made, will place constraints on the depth of the very peculiar ``Stealth'' region of Mars and on the thickness of polar ices on Mars and Mercury could be obtained. While it may seem that Magellan answered all of the questions regarding the surface of Venus, it must be remembered that the radar on Magellan was a single polarization (1 linear), single frequency (2.4 GHz) instrument. A 2.4 GHz system at the VLA would make good polarized measurements of the surface of Venus possible which would greatly enhance the usefulness of the radar data from the Magellan mission. Because of attenuation by the Venusian atmosphere, 8.4 GHz data is severely compromised. The lack of a 2.4 GHz system at the VLA is the one factor preventing very fruitful collaborations with both Arecibo and Goldstone in this area.
Second, in order to support the Cassini mission to Saturn, JPL is committed to installing a Ka-band transmitter. Many technical issues are still not settled, but this may provide another radar frequency at GHz. The power will be much lower than available 2.4 and 8.4 GHz radar systems ( KW). However, taking into account the theoretical gain of a telescope, the scaling of the return width with frequency and the increased bandwidth of reception, it appears radar between Goldstone and the enhanced VLA in the 33 GHz band would produce a signal to noise ratio of about 1/4 of the current Goldstone/VLA 8.4 GHz system. Therefore, bistatic radar in the 33 GHz band will provide for higher resolution studies of selected objects and extend the frequency coverage available to study surface properties by almost a factor of four.
Third, the new correlator will be of enormous benefit to planetary radar studies in several respects. Because different portions of a rotating planetary surface are approaching the radar line of sight at different velocities, the received radar power is spread in frequency by an amount which depends on the rotational period. For example, in the 8.4 GHz band, the spectral broadening varies from kHz for Mars to about Hz for Venus. Typical bandwidths for asteroids are on the order of 100 Hz at 8.4 GHz. The frequency spreads scale linearly with wavelength, so widths in the 2.4 GHz band are about a factor of 4 smaller. With the current system at the VLA, the smallest channel width available is 380 Hz for a single IF. Only the largest of the asteroids are resolved and only if data from one polarization are recorded. For a typical Goldstone/VLA radar experiment, unless one is willing to give up the polarization information, the minimum channel width is 760 Hz. Unfortunately, this is wider than all of the returned signals except for Mars and barely comparable to those from the icy satellites. Because of inadequate resolution, Doppler information is lost and in addition, a penalty is paid in signal to noise ratio, since in those frequencies where there is no radar echo power, there is still noise power. In fact, in order to maximize the signal to noise on a particular surface feature, a channel width equal to the spectral width of that feature is desirable.
The loss of Doppler information limits the experiments in other ways as well. In the case of Titan, such information could be used to calculate the position of the rotational pole, about which almost nothing is known. Determination of the rotational parameters of asteroids could also be done throughout the main belt with the upgraded Arecibo 2.4 GHz transmitter. For these reasons, much higher frequency resolution at the VLA is desired. This is only obtainable with a new correlator with a frequency resolution Hz for all of the bodies except Mars, for which resolution on the order of 100 Hz would be adequate.
In all of these radar experiments, there are two quantities which are currently not measured which would be of much use. The first is the amount of thermal emission contributed to the signal. In some cases this is desired because of the extra information it provides (knowledge of reflectivity and emissivity at the same frequency ties down the surface properties much better), but in others it is essential to interpret the radar data. For objects that contribute significant thermal flux (planets) to the observed signal, some attempt must be made to estimate the thermal emission flux and subtract it out. Currently, that flux is estimated from spectral channels outside of the bandwidth of the received echo flux. However, because work to date has been constrained to be in narrow spectral channels, the total bandwidth of the channels from which thermal emission is estimated is itself very narrow.
The other quantity which is currently not obtainable from VLA/Goldstone radar experiments is the total power (zero-spacing flux), a quantity which is particularly important for planetary observations. These experiments are generally performed in the A configuration to get the highest angular resolution. The price is a serious short spacing problem. This would be alleviated with a total power measurement system. It should be accurate to a few percent, but for our purposes it does not need to be measured often (once every minutes would be sufficient).
Finally, in order to achieve angular resolution at 2.4 GHz sufficient to permit direct comparisons with 8.4 GHz results, baselines which are about a factor of 4 longer than the longest currently available at the VLA are required. In addition, on these baselines, the polar caps of Mars could also be resolved at 8.4 GHz. The addition of the innermost 4 VLBA antennas (PT, LA, KP, FD) and several new VLA/VLBA antennas would fulfill this need.