Radar Imaging of Near Earth Asteroids

Radar is already the most powerful ground-based technique for post-discovery investigation of NEOs (near-Earth objects). Radar is uniquely capable of resolving NEOs spatially by measuring the distribution of echo power in time delay (range) and Doppler frequency (line-of-sight velocity) with very fine precision in each coordinate. Current capabilities utilizing the Goldstone, 70 m NASA Deep Space Network antenna, and the Arecibo, 305 m telescope, are for best resolutions of order 10 m in range and 0.1 mm/s in velocity. The fractional precision of delay-Doppler positional measurements plus their orthogonality to optical astrometry makes them invaluable for refining orbits and prediction ephemerides. A single radar detection secures the orbit well enough to prevent ``loss'' of the object, shrinking the object's instantaneous positional uncertainties by orders of magnitude with respect to an optical only orbit and greatly improving the accuracy of long-term trajectory predictions. Additional advantages come from radar wavelength's sensitivity to near surface bulk density and roughness. For comets, radar waves can penetrate optically opaque comas to examine the nucleus and can disclose the presence of macroscopic coma particles.

One of the more spectacular NEO results of recent years has been a sequence of delay-Doppler images of 4769 Castalia (1989 PB) shown in Fig. 3.2 (Ostro et al. 1990) obtained two weeks after its Aug. 1989 discovery. These reveal it to consist of two kilometer-sized lobes in contact. Least-squares estimation of Castalia's three-dimensional shape from the radar images supports the hypothesis that Castalia is a contact-binary asteroid formed from the gentle collision of the two lobes and also constrains the object's surface morphology and pole direction.

  

Figure 3.4: Radar images of asteroid 4769 Castalia (1989PB). This 64 frame Arecibo movie is to be read like a book (left to right in the top row, etc.). The radar lies toward the top of the figure in the image plane, which probably is about 35$^\circ$ from the asteroid's equator. In each frame, echo power (i.e., the brightness seen by the radar) is plotted vs. time delay (increasing from top to bottom) and Doppler frequency (increasing from left to right). The radar illumination comes from the top of the figure, so parts of the asteroid facing toward the bottom are not seen in these images. The object, each of whose lobes is about a kilometer in diameter, is seen rotating through about 220$^\circ$ during the 2.5 h sequence. These images have been smoothed from the original resolution of 2 $\mu$s $\times$ 1 Hz (0.3 $\times$ 0.17 km). (from Ostro et al. 1990).

Another illustrative result is the delay-Doppler imaging of 4179 Toutatis in Dec. 1992 (Ostro et al. 1995). Resolutions as fine as 125 ns (19 m in range) and 8.3 mHz (0.15 mm/s in radial velocity) were obtained, placing thousands of pixels across the asteroid. The images shown in Fig. 3.5 reveal this asteroid to be in a highly unusual, nonprincipal -axis (NPA) spin state with several-day characteristic time scales. This data set provides physical and dynamical information that is unprecedented for an Earth-crossing object.

  

Figure 3.5: Radar images of asteroid 4179 Toutatis from 1992. Goldstone low-resolution images from Dec. 2-18 (top three rows) and Arecibo images from Dec. 14-19 (bottom row) are plotted with time delay increasing toward the bottom and Doppler frequency increasing toward the left. On the vertical sides, ticks are 2 $\mu$s apart. Two horizontal sides have ticks separated by 1 Hz for Goldstone and 0.28 Hz for Arecibo; those intervals correspond to a radial velocity difference of 18 mm/s. (from Ostro et al. 1995).

Use of the SKA in a bistatic radar configuration with Arecibo or Goldstone as the transmitter would have a major impact on the study and detection of NEOs. The factor of 30-200 increase in receiver sensitivity (with Arecibo and Golstone respectively) would permit objects of smaller diameter to be observed out to substantially larger distances. The number of objects accessible to radar study would therefore increase dramatically. In addition, the high angular resolution of SKA (as high as 5 mas) would permit simultaneous spatial imaging of the illuminated surface, greatly simplifying the physical modeling. The linear resolution afforded by such a spatial resolution, 400 m at a distance of 0.1 AU, is such that a detailed three dimensional image of each object could be obtained instantaneously. A single observing session would then also provide extremely high precision in the orbital determination, since the three dimensional space velocity would follow from the proper motion observed in an interval of only a few hours.