Geoscience Reference
In-Depth Information
penetration of the radar energy into a material tends to
increase with longer wavelength (and 'ground penetrating
radar' (GPR) used in field measurements tends to use longer
wavelengths than imaging radars—often about 1 m).
Generally, sand dunes look somewhat dark to radar
because dunes are usually smooth on the scale of the radar
wavelength and thus specularly reflect most of the energy
away from the transmitter, and the dune has little internal
structure to scatter any energy that does enter the dune.
Thus apart from topographic glints, dunes generally appear
dark against what is usually a rougher and/or denser in-
terdune, and indeed, on the global scale, the sand seas are
the darkest land areas (see Fig. 18.19 ). However, brightness
on the dunes themselves can be highly variable, with strong
glints from slip faces that are oriented towards the radar (see
Fig. 18.23 ).
Whether truly specular glints are present or not, the
backscatter curve (e.g., Fig. 18.20 ) can be exploited to
gauge the average orientation of the surface within the
pixel. This approach, termed radarclinometry as an analog
to photoclinometry, works best on surfaces that have a
uniform composition. With a known viewing geometry and
a backscatter curve assumed to be uniform, the slope and
thus the height of resolved dunes can be calculated. This
was done on Titan by Lorenz et al. (2006), determining
heights of over 100 m: Neish et al. (2010) performed further
measurements on Titan, and also showed that the method
works on terrestrial dunes (in the Namib—Fig. 18.24 ) and
is not significantly degraded by the lower resolution of the
Cassini radar compared with terrestrial radars.
Measuring dune size, spacing and orientation is of course
as easy with a radar image as with an optical one. Because
radar engineers are familiar with working in the frequency
domain, it has been natural to apply spectral methods
(Fourier transforms) on the image (e.g., Qong 2000; such
approaches are more routinely applied to SAR images of the
sea surface, to determine wavelength and orientation of
waves, e.g., associated with hurricanes, so it is a straight-
forward step to apply the same algorithm to dunes).
However, unlike (e.g., polar terrain) dunes tend to be
found in dry deserts, where optical visibility is good and
thus at the spatial scale afforded by radar imaging, the
terrestrial sand seas had already been mapped optically
(e.g., McKee et al. 1979). While radar has been the prin-
cipal tool in discovering and characterizing dunes on Titan
and Venus (which are optically very challenging), most of
the terrestrial literature has been devoted to understanding
how dunes appear to radar (i.e., assessing the role of dif-
ferent contributions to the echo) rather than to actually
discovering something new about the terrestrial dunes.
Some likely potential applications are to estimate the small-
scale slope distributions over wide areas, or to estimate the
thickness of interdune sand deposits.
Fig. 18.20 Brightness of a horizontal surface illuminated by radar as
a function of roughness and incidence angle (0 incidence is vertical).
The effective incidence of an arbitrary surface will also depend on its
slope relative to the illumination direction—this sensitivity is partic-
ularly strong at low incidence angles, making radarclinometry possible
and leading to glints. At shallower incidence angles, the wavelength
scale roughness is usually the dominant factor. (Optical reflectivity
shows similar behavior, but most geological surfaces are optically
rough.)
realized. There are a number of good texts on radar for
planetary surface observation, notably Elachi (1988); the
three-volume epic by Ulaby et al. (1982) is widely regarded
as the classic text for serious radar scholars. A short but
useful guide was produced for the Magellan mission (Ford
et al. 1993) and is available online.
For now, we will discuss radar reflectivity and the
interpretation of images in general terms (and Fig. 18.22
summarizes this). The relation of reflectivity to slope and
roughness is adequate description for many terrestrial sur-
faces where the radio energy is simply reflected at the
surface, especially where water is present. However, radar
backscatter can be complicated somewhat by the penetra-
tion of radar energy into a target and its subsequent
reflection within the material, termed 'volume scattering'.
This can occur in materials that are free of moisture (which
includes very cold ice), and in fact is quite common for
porous, dry materials like desert sands, and for the lunar
regolith. Radar can often see 'through' sand sheets with
thicknesses of tens of centimeters, and the SIR experiments
revealed, for example, river channels in bedrock that had
been
buried
by
more
recent
aeolian
deposits.
The
 
 
Search WWH ::




Custom Search