Remote Sensing - Dune Worlds: How Windblown Sand Shapes Planetary Landscapes - page 212

Originally, such change detection measurements
required careful matching of images (perhaps via wet
photographic processing) and many tedious manual mea-
surements. Now computational tools such as Geographical
Information System (GIS) products, Google Earth and so on
make such measurements easier. Furthermore, automated
and semiautomated processes can now extract differences in
position of features by a fraction of a pixel, by computa-
tionally cross-correlating images. One, but by no means the
only, implementation of this technique is called COSI-
CORR (Co-registration of Optically Sensed Images and
Correlation), developed at Caltech. Similar automated cor-
relation methods are used (with different viewing geome-
tries) to make digital elevation models from stereo imaging.
It may be remarked that the hand-held image is not dead:
high-quality digital stills and video are routinely obtained
by astronauts on the International Space Station (there is
even a big high-quality windowed cupola for Earth obser-
vations), and NASA makes these often striking data avail-
able too (and indeed some of the illustrations in this topic
come from this site). The cameras used here are typically
true-color and, of course, because someone thought the
image was worth pressing the shutter button for, are often
well-composed and appealing. Notably, many images are
off-nadir (e.g., Fig. 18.7 ), which makes a pleasing change in
perspective from the flat view normally used in Earth
observation.

Fig. 18.8 Surface temperature behavior of Mars as a function of
thermal inertia and albedo. Top, diurnal temperatures for a range of
thermal inertia during the summer, assuming an albedo of 0.10.
Generally, thermal inertia controls the amplitude and the phase of the
diurnal temperature cycle and albedo controls the mean. Mars hour
represents 1/24th of a Mars solar day (88775 s). After Fig. 18.1 of
Mellon et al. (2008)

are found can look blueish, whereas to the human eye they
would be simply white (the same effect is seen on some
Mars images).
Until a decade or so ago, resolutions much better
than *20 m for scientific (as opposed to military surveil-
lance) applications meant using aerial photography. How-
ever, electronic cameras have dramatically improved, as
well as the capability for relatively affordable spacecraft to
achieve the pointing accuracy needed to image targets on-
demand, and (importantly) the connection of customers
with providers made not just possible, but easy, by the
internet. As a result, a variety of commercial enterprises
worldwide now provide satellite imaging down to less than
0.5 m resolution, which can be explored easily with tools
like Google Earth.
Geomorphologically, the Earth's dunes are overall well-
characterized globally at the 'free' 20 m level (e.g.,
Figs. 18.5 and 18.6 , from ASTER and Landsat). However,
the higher spatial resolution now available commercially
means that areas of particular interest, notably those where
dunes are migrating, can be studied much more closely
without mounting an expensive or hazardous expedition
(both a good and a bad thing!). This higher resolution
permits changes to be detected over much shorter time-
scales than was the case in the past (e.g., only the fastest-
moving barchans were clocked by Landsat data). We report
our own analysis of barchan migration using commercial
imaging of the Star Wars film set in Tunisia in Chap. 24 .

18.3 Thermal Imaging

In addition to imaging reflected light, it is possible to sense
thermal infrared light emitted from surfaces. This gives
indications of surface temperature, as well as being some-
what influenced by composition. The surface temperature of
course varies by latitude, season and time of day, as well as
factors such as the particle size, and whether volatiles are
present.
A prominent technique in planetary remote sensing is to
compare daytime and nighttime thermal images. Rocks and
hard ground warm up and cool down more slowly (an effect
due partly to density q and heat capacity c, and in part due
to thermal conductivity k) and so are cooler in the day and
warmer at night than are porous materials like sand and
dust. Table 18.1 lists the thermal conductivity for a few
rock and sand examples, values that are up to two orders of
magnitude lower than the thermal conductivity of metals.
Density represents the mass per unit volume of the material.
The table illustrates that density does not vary greatly for
most natural materials; water is the least dense common
material.

Heat capacity (also called the specific heat