Introduction to Planetary Radar

Arecibo Observatory

Since the completion of the Arecibo upgrade, the new radar system has been used for a wide range of solar system studies. Radar images reveal a wealth of information about the shapes and surface properties of solid bodies in the solar system. The Arecibo telescope has collected data on Mercury, Venus, Jovian satellites, and Saturn's rings and satellites, and numerous asteroids and comets. Some recent results are summarized in these figures.

A major component of the upgraded telescope is the installation of a new 1 megawatt 12.6 cm radar system, which has been jointly funded by NASA and the NSF. This instrument will enable the Observatory to detect asteroids over a very much wider radial range than with the old 420 kilowatt system. We anticipate as much as a factor 40 improvement in sensitivity compared to before the Gregorian Upgrade (and in no case less than a factor of 14). Some of the improvement comes from the increase in transmitter power, some from the optics of the telescope, and some from the greatly increased zenith angle range with constant sensitivity that the new ground screen provides.

Available equipment includes a 12.6 cm (2380 MHz) dual-polarization maser receiver with the capability of both complex voltage sampling and a new, high-throughput, radar decoder. Sampling has a limit of 10 MHz, while the decoder can operate at up to 20 MHz (7.5m range resolution).


One of the major recent successes in the field of planetary radar astronomy was the discovery and mapping of deposits of volatiles on the shadowed floors of impact craters at the poles of Mercury. Because of the similarity of their radar scattering properties to those of icy surfaces in the solar system, the deposits are thought to be water ice. The figure shows the most recent Arecibo 12.6 cm wavelength radar image of the north polar region of Mercury made at a resolution of 1.5 km by J. Harmon (NAIC), P.Perrilat (NAIC) and M. Slade (JPL). It was published in the January, 2001 issue of Icarus. The donut shape of the deposits close to the pole are due to the presence of central peaks in the craters while the bright arcs away from the poles coincide with the shadowed areas for these craters. A detection of the "ice" at the north pole was also obtained with the Arecibo 70 cm wavelength radar indicating that the deposits may be at least several meters thick.
Mercury North Pole

Arecibo S-band radar image of the north polar region of Mercury by J. Harmon, P. Perrilat, and M. Slade. The resolution is 1.5 km and the image measures 450 km on a side. The bright features are thought to be ice deposits on permanently shadowed crater floors.


Radar images of Venus were obtained using the Arecibo radar system in August, 1999, and again in March/April, 2001. The 2001 observations included the first scheduled use of the new Robert C. Byrd 100-meter Green Bank Telescope. The Arecibo Observatory and the GBT together serve as a radar interferometer with the objective of creating high resolution topographic maps of Maxwell Montes at spatial resolutions adequate to determine the relationship between areas of high reflectivity/low emissivity and altitude. This work has been carried out by D. Campbell (Cornell/NAIC), J.-L. Margot (Caltech), L. Carter (Cornell), B. Campbell (NASM) and J. Dorris (Cornell). Other objectives include looking for surface features which had changed since the Magellan mission, and to obtain images in the four Stokes parameters to look at the polarization properties of the surface. Of particular interest are the floors and ejecta deposits of impact craters and the high reflectivity/low emissivity surfaces found at high elevations. Linear polarization of 10 to 40 percent is found in several areas of the maps. Areas with high linear polarization seem to be associated with radar-bright regions near impact craters.

The delay-Doppler images below were obtained by transmitting continuously from Arecibo Observatory and receiving the radar echoes at the 100 meter Green Bank Telescope. This SC image (same sense circular polarization as that transmitted) of the front part of Venus has been spatially averaged to approximately 5 km resolution.

This SC image shows a portion of Maxwell Montes, spatially averaged to 1.2 km resolution.

The Moon

The discovery of ice at the poles of Mercury triggered a search for ice in the shadowed floors of craters near the lunar poles as part of Cornell graduate student, Nick Stacy's Ph.D. thesis work. Stacy developed sophisticated data acquisition and analysis techniques to image selected regions of the Moon in all four Stokes' polarization parameters at resolutions down to 20 m using the 430 MHz radar system. No clear evidence of ice was found, although there are a small number of areas of anomalous backscatter at the south pole, which warrant further study. These lunar observations have, for the first time, demonstrated the capability for radar measurements to map, assuming a scattering model for the surface, variations in mare surface dielectric constant and areas with high concentrations of titanium or iron oxides.
Moon South Pole
This radar image in delay-Doppler coordinates of the south pole region of the Moon was made at Arecibo in a search for ice in permanently shadowed areas. It is 400 km in each coordinate and the original image has a resolution of 500 m in delay (vertical) and 580 m in Doppler (horizontal). The illumination is from the top (so range increases downward), and increasing Doppler frequency is towards the left. The approximate location of the south pole is indicated by the cross. The search was for the characteristic radar signature of ice, high backscatter cross section and high circular polarization ratio. No clear indication of ice was found, although a small number of areas with anomalous radar properties need further investigation. These lunar investigations were done by N. J. Stacy as part of his Cornell Ph.D. thesis.


Radar imaging of near-Earth asteroids can provide dramatic images with resolutions down to 8 m, comparable to the images obtained by the Galileo and NEAR-Shoemaker spacecraft. The rotation rate, shape and reflectivity give us information about the asteroids' density, internal structure, and surface properties. The images also show surface features such as impact craters, and irregularities, which can often be traced across the surface as the object rotates. A series of images can be used to derive a three-dimensional shape model. In the case of 216 Kleopatra, shown below, that shape can be highly irregular, and is a strong constraint on formation mechanisms. The shape of some very small objects is surprisingly spherical, which suggests a rubble pile with no internal strength. Two near-Earth binary asteroid systems have now been discovered by radar. Both 2000 DP107 and 2000 UG11 were determined to be binary systems from a combination of Goldstone and Arecibo radar observations. These close pairs of asteroids must be recently formed, perhaps by tidal forces on a previous close approach to the Earth. The lifetime of such a binary system against collisional disruption is quite short. Main-belt asteroids have also been studied, and images with resolution comparable to those obtained by the Hubble Space Telescope have been obtained.
Binary asteroid 2001 UG11
Asteroid 2000 UG11 is the second near-Earth binary asteroid known. The primary, in the upper part of the image, is about 230m diameter. The secondary, the bright dot below, is about 100m, and separated by at least 300m from the primary. The orbital period is 19.5 +/- 1.5 hours. This image of 2000 UG11 was obtained on 2000 Nov 7. Range increases downwards, delay increases to the left. The range resolution is 15m per pixel. The cross-range resolution is approximately the same, but depends on the viewing geometry, which is not yet completly determined.
Asteroid 2001 GQ2

Asteroid 2001 GQ2 was imaged by the Arecibo radar system during April, 2001 by M. Nolan. This sequence of images shows the asteroid appearance during about 1.5 hours on 29 April. The first image is in the upper left, with time increasing across and then down. Each image is a sum spanning about 12 minutes. The image resolution is about 15m, and the asteroid is estimated to be 400m in diameter. In each image, range increases to the right, and doppler frequency increases downward. CCD images obtained by M. Nicolini, using a 0.4m telescope at the Observatory of Cavezzo, showed that the rotation period was longer than 8 hours (no maximum or minumum in 2 hours). The radar data confirms that the rotation period is probably about 10 hours.
Asteroid 216 Kleopatra

These radar images of main-belt asteroid 216 Kleopatra were obtained in November, 1999, about one hour apart (Ostro et al 2000, Science). The top row are the data, which are delay-Doppler images with range increasing down, frequency to the right. The middle row are simulated radar data, using the shape model derived. The bottom row shows the shape model, as it would appear illuminated in the plane of the sky. The model was derived from 12 images on four days, together with a diameter from an earlier stellar occulation. The size bar in the lower left is 100 km.

Other interesting Asteroid radar links:

Steve Ostro's asteroid radar page

Scott Hudson's asteroid page (mostly radar-based)


Comet C/2001 A2 (LINEAR) was discovered as part of the Lincoln Laboratory Near-Earth asteroid survey (LINEAR) on 16 January 2001. Observing conditions were favorable at Arecibo July 5-15 2001, and observing time for both spectroscopy of 18-cm OH lines and S-band radar observations were arranged. The comet nucleus was observed to split into two pieces on April 30. Component A broke into two pieces May 16, and faded away within a week. Component B, now C/2001 A2-B LINEAR, continued with nearly the original predicted brightness. Three additional fragments were seen to separate from A2-B in early June, at which time there was an increase in gas and dust emission.

The Arecibo observations of the OH lines took place 5-6 July 2001. The OH in the cometary coma absorbs solar radiation, which can be re-emitted as microwaves under favorable conditions. The emission strength depends on the radial heliocentric velocity of the comet and on the amount of OH in the coma. The 1612, 1665, and 1667 MHz OH lines were observed, but only the 1667 MHz line was detected. In addition to looking at the nucleus, we also observed one beam width away (4.1 arcmin), in a hexagonal pattern around the nucleus, oriented along the sun-tailward direction. The figure shows the nucleus spectrum and the sum of all six of the offset positions. The strengths of these lines will be compared to model predictions to determine the outflow velocity, identify asymmetries, and to search for signs of collisional quenching. In very active comets, such as Hale-Bopp, the OH lines were thermalized or ``quenched'' in the inner 500,000 km of the coma. One of the goals of this investigation is to determine the extent of quenching in more ``normal'' comets with lower gas production rates.

The spectrum of Comet LINEAR A2-B shows the 1667 MHz OH line in emission.

Because the comet passed within 0.25 AU of the Earth in early July, when it was visible from Arecibo Observatory, it was fairly well suited to radar observation. We detected the comet with the S-band (2380 MHz) radar in continuous wave (CW) mode on July 7-10. The spectrum shows a broad profile of return echo power from the cometary coma. There is no clear central spike, indicating return from the nucleus, contrary to what has been seen on other comets (Harmon et al 1999). Some return was also seen in the unexpected or same-sense polarization, at 20% of the expected sense. This indicates that the particles in the coma are comparable in size to the radar wavelength (12.6 cm). Given the fragmentation of this comet in the previous months, it may not be surprising to find large-sized debris in the coma. The gas and dust production from this comet is close to the prediction for a several-km nucleus. We attempted to image the comet on July 9 and 10, and those data are still being analyzed.

This CW echo from Comet LINEAR A2-B shows a broad return from the coma, in both opposite-circular (upper plot) and same-circular (lower plot) polarization. This indicates that the particles are comparable in size to the radar wavelength (12.6 cm) No clear return from the nucleus is seen.

Jupiter's satellites

A program of delay-Doppler mapping of the Galillean satellites is underway by L. Harcke and colleagues at Stanford University. The surfaces of Europa, Ganymede and Callisto have been imaged in order to determine the surface properties and look for different radar properties of the various terrain types. These maps will be correlated with the Galileo images of the satellites.

This image is a SC radar albedo map of the western hemisphere of Callisto. An ambiguity exists between the northern and southern halves of the planet due to the range-Doppler imaging technique employed, thus the image appears symmetric about the equator. A projection effect causes the resolution cells to be long and thin nearer to the equator. Signal to noise ratio is strongest near the sub-Radar point in the center of the image, and decreases moving towards the poles and the limb. The feature marked with a red X is several standard deviations above the noise level. A study correlating this object with labeled features in the Galileo mission optical image archive is ongoing.

Saturn's rings

In fall of 1999 and 2000, the rings of Saturn were imaged using the Arecibo S-band radar system by P. Nicholson, D. Campbell, R. French, G. Black and J.-L. Margot. The image shown is a sum of 5 days of dual-circular polarization data, co-adding both polarizations. The Keplerian velocity profile of the rings results in the outermost or A ring which provides the earliest echo, appearing at a lower Doppler shift than the middle or B ring. This leads to four bright ``crossover'' regions on either side where signal from the two different rings add together, analagous to the north/south ambiguity for radar imaging of rigidly rotating bodies. A pronounced azimuthal asymmetry can be seen: the rings are brighter on the far quadrant on the receding (western) ansa than on the near quadrant of the approaching (eastern) ansa. The most widely accepted explanation of the asymmetry involves gravitational ``wakes'' generated by individual large ring particles, which are distorted by Keplerian shear into elongated structures trailing at angles of about 70 degrees from the radial direction (e.g. Franklin and Colombo 1978).
Saturn's Rings

A delay-Doppler image of Saturn's rings at a frequency of 2380 MHz (12.6 cm) is compared to a model image constructed by reprojecting a pair of HST images taken at 439 nm. Time delay increases from bottom to top, and Doppler shift increases from left to right. The effective spatial resolution is 2000 km by 15000 km.

Questions or comments? Contact Ellen Howell. (

Last modified 2001 Aug 24