Chapter 5 EO—Spectral Imagery

Chapter 5 EO—Spectral Imagery
Chapter 5
EO—Spectral Imagery
The discussion of remote sensing up to this point has focused on panchromatic
(black & white) imagery. Beyond recording obvious features of size and shape, remote sensing excels in capturing and interpreting color. Color systems also yield
some spectacular imagery. For example, in this early “true-color” image from
Landsat 7, we perceive the green hillsides and muddy runoff of the upper San Francisco bay (Fig. 5.1).
5.1 Reflectance of Materials
The reflectance of most materials varies with wavelength. This allows spectral
imagers, such as those on the Landsat missions, to distinguish different materials.
Distinguishing between minerals is a fairly common goal for such work by geologists.
In Fig. 5.2, different aspects of reflective spectra are illustrated. Spectra are the
fingerprints of elements, deriving from their fundamental atomic characteristics, as
indicated in our discussion of Bohr’s model of the hydrogen atom. One of the more
important, and dramatic, spectral features found in remote sensing is the “IR ledge”
at 0.7 µm, as found in Fig. 5.2.1 This dramatic rise in reflectance with wavelength
makes vegetation appear bright in the infrared (old-style, black-and-white infrared
film would show vegetation as white, for example). The military designs camouflage to mimic this behavior. The panchromatic sensors on Landsat, SPOT,
IKONOS, and Quickbird extend well into the infrared, and as a result, vegetation is
bright in their imagery.
Also occasionally termed the “red edge,” e.g., in D.N.H. Horler, M. Dockray, and J. Barber,
“The red edge of plant leaf reflectance,” Int. J. Remote Sens. 4, pp. 273–288 (1983).
Chapter 5
Figure 5.1 Visible image of San Francisco from Landsat 7, April 23, 1999, flight day 9, orbit
117. ~1830Z With special thanks to Rebecca Farr, NESDIS/NOAA. The satellite is not yet in
its final orbit, and not on the standard reference grid, WRS, so the scene is offset 31.9 km
east of nominal scene center (Path 44, Row 34). Landsat has been the premier Earth resources satellite system for three decades.
5.2 Human Visual Response2
Before considering the spectral response of orbital systems, let’s consider human
visual response. The sensitive elements of the eye are the rods and cones. Rods
(which far outnumber cones) are sensitive to differences in brightness within the
See W.L. Teng, E.R. Loew, D.I. Ross, V.G. Zsilinsky, C. Lo, W.R. Philipson, W.D. Philpot,
S.A. Morain, “Fundamentals of Photographic Interpretation,” 2nd ed. in Manual of Photographic Interpretation. W.R. Philipson (Ed.), American Society for Photogrammetry and Remote Sensing,
Bethseda, MD, p. 67 (1997). Also see H.J.A. Dartnall, J.K. Bowmaker, and J.D. Mollon,
“Microspectrophotogrammetry of Human Photorecepters,” in Color Vision: Physiology &
Psychophysics, John Mollon, L. Ted Sharpe (Eds.), Academic Press, London (1983).
EO—Spectral Imagery
Figure 5.2 Reflectance for some manmade and natural materials. Note how the olive-green
paint mimics the grass spectrum in the visible to NIR, but then deviates.
Figure 5.3 Human vision: The white curves indicate the sensitivity level for the three types of
cones. The black curve indicates the sensitivity of the rods.3
middle of the light spectrum. The rods’ peak sensitivity corresponds to the peak in
solar illumination. If we had only rods, we would see in shades of grey. 3
See John E. Dowling, The Retina: An Approachable Part of the Brain, Harvard University Press,
Cambridge, Mass., (1987). See also H.J.A. Dartnall, J.K. Bowmaker, and J.D. Mollon, “Human Visual Pigments: Microspectrophotometric Results from the Eyes of Seven Persons,” Proceedings of
the Royal Society of London B, 220 (1218), pp. 115–130 (1983).
Chapter 5
Cones provide our color vision. There are three types of cones:
L-cones are sensitive primarily to red in the visible spectrum.
M-cones are sensitive to green.
S-cones are sensitive to blue.
5.3 Landsat
In late July 1972, NASA launched the first Earth-Resources Technology Satellite,
ERTS-1. The name of the satellite, and those that followed, was soon changed to Landsat. These have been the primary Earth-resources satellites ever since, utilizing
multispectral imagery (MSI) with a spatial resolution of 30 m. The most recent in the
series, and perhaps the last, is Landsat 7. The table below shows some of the parameters
for the sequence of missions. Note the evolution in data storage technology, bandwidth, and the changes in downlink technology. Resolution has gradually increased
with time. Landsat 7 added a 15-m GSD panchromatic sensor to the MSI sensors.
Figure 5.4 shows the satellite orbit, and illustrates the 185-km viewing swath
for the Enhanced Thematic Mapper (ETM) sensor, described below. The figure
also reflects the different downlink options, as the sensor communicates with the
Landsat Ground Station (LGS).
Table 5.1 Landsat sensors and orbits.
Landsat 1
Landsat 2
Landsat 3
July 23, 1972–
January 6, 1978
January 22,
25, 1982
March 5, 1978–
March 31, 1983
July 16, 1982–*
Landsat 4
Landsat 5** March 1, 1984– MSS
to date
Landsat 7
April 15, 1999–to ETM+
Data Link
(85 Mbps)
(15 Mbps)
w/solid state
(150 Mbps)
* TM data transmission failed in August 1993; satellite was used for maneuver testing.
** Landsat 6 failed due to a launching problem.
*** RBV: Return Beam Vidicon; MSS: Multi-Spectral Scanner; TM: Thematic Mapper
EO—Spectral Imagery
Figure 5.4 The nadir-viewing satellite
images a swath 185 km wide.
Figure 5.5 Subsequent orbits are displaced by some 2500 km to the west.
There are 233 unique orbit tracks.
5.3.1 Orbit
The Landsat missions have been sun-synchronous polar orbiters in circular orbits
(LEO). The later missions operating at an altitude of 705 km. NASA’s Mission to
Planet Earth has added several satellites in the Landsat 7 orbit—Terra, Aqua,
SAC-C, and EO-1. They trail the older satellite by a few minutes in orbital sequence. An inclination of ~97° makes them sun-synchronous. Equatorial crossings
were set at 9:30 am for Landsats 1, 2, and 3; 10:30 am for Landsats 4 and 5; and
10:00 am for Landsat 7. Landsats 4, 5, and 6 have orbit tracks set for 14.5 orbits per
day. The repeat cycle is every sixteen days. Figure 5.5 shows the orbit track, and
shows how the orbit shifts in local time (longitude).
The orbit track for Landsat 7 is further illustrated in Fig. 5.6. The ground track
is illustrated for 2 orbits. The satellite is on an ascending node on the night side, and
descends southward on the day side.
Chapter 5
Figure 5.6 This orbit ground track corresponds to the San Francisco image above. The yellow spot just below Mexico City is the sub-solar point, April 23, 1999, ~1830Z.
5.3.2 Sensor: Thematic Mapper4
The first Landsats flew with a multispectral (MSS) imager and a TV-like sensor
(the return-beam vidicon, or RBV). Beginning with Landsat 4, the primary instrument became the Thematic Mapper (TM), though the MSS was also carried to
maintain continuity in archival, synoptic datasets. The TM sensor has provided
seven bands of spectral information at 30-m resolution since 1982. On the most recent vehicle, Landsat 7, the instrument was revised as the Enhanced Thematic
Mapper plus, or ETM+. The instrument features improved spatial resolution in the
LWIR channel (60 m) and a new panchromatic band with higher spatial resolution
(15 m). The discussion that follows addresses the ETM+ sensor. Optics
The optical design of Landsat predates large linear or rectangular arrays; it is of
whiskbroom design. The telescope is a Ritchey-Chrétien Cassegrain, as seen with
several earlier systems. The primary mirror (outer) aperture is 40.64 cm; the clear
inner aperture is 16.66 cm. The effective focal length is 2.438 m, and the aperture is
f/6. The instantaneous field of view (IFOV) for one pixel of the high-resolution
panchromatic sensor is 42.5 µrad.
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF