IEEE Geoscience and Remote Sensing Magazine - June 2019 - 76

for airborne off-nadir (long slant-paths) and ground standoff geometries. However, the contribution of the various
terms is greatly influenced by its geometry (see Table 1). It
is sometimes assumed that
standoff data are easier to
analyze and interpret than
MINERALS HAVE STRONG
airborne data. In general, this
SPECTRAL CONTRAST WITH
is not true, because standoff
SOME FEATURES THAT
data have a very strong contribution from the reflected
APPEAR AS EMISSIVITY
downwelling radiance. ReMINIMA (REFLECTANCE
flected downwelling radiance
MAXIMA) IN THE LWIR
can cause apparent spectral
REGION.
features in the retrieved surface emissivity signature,
which degrades interpretation
capabilities and may also cause misidentification. This is
particularly difficult to manage for nondiffuse targets measured at high emission angles [7].

CHALLENGES
Many of the data analysis techniques developed for the reflective domain can be employed for LWIR data; however,
the interpretation of the results is more challenging because
of the many physical effects in play. An example is the small
but important reflection of downwelling radiance. Materials that have an appreciable reflectance (low emissivity) in
the LWIR region will exhibit significantly different ground
radiance in the presence of clouds than they will under a
clear sky because clouds provide substantial downwelling
radiance in the IR. This is especially true in the MWIR region, where most materials have emissivity between 0.5
and 1. Additional challenges caused by the coupling between temperature and emissivity are discussed in subsequent sections.

Emissivity (Offset for Clarity)

IMAGING SPECTROMETERS
The optical principles that provide the spectral discrimination of an imaging spectrometer are the same regardless of
the spectral range. Instruments can
be divided into two broad categories: those that discriminate between
wavelengths based upon the optical
1
Gypsum
properties of a material and those
0.9
that use the wave nature of light.
Quartz
The former employ filters or prisms,
0.8
Alunite
and the latter utilize diffraction and
0.5
interference [1]. Instruments from
0.9
both categories have been widely
1
used for LWIR imaging spectromAlbite
0.9
0.7
eters. The latter type can be further
Kaolinite
divided into designs that employ
0.9
0.95
gratings (where the wavefront is spa1
Montmorillonite
Muscovite
tially separated as a function of the
wavelength) and those that employ
0.6
beam splitters for amplitude division
0.95
0.8
Actinolite
of the wavefront. This latter category
includes both Michelson and FabOpal
0.85
0.9
ry-Pérot interferometers [8]. The
1
0.8
different types are illustrated in
Hedenbergite
Calcite
Figure 5. Both dispersive and inter0.9
0.9
ferometric optical forms have been
1
0.8
perfected and are highly efficient for
Fayalite
the currently available focal plane ar1
Vegetation
ray (FPA) sizes and pixel pitches. The
0.8
0.9
standard forms will be addressed in
both cases. Additional forms have
8
9
10
11
12
8
9
10
11
12
been proposed, such as linear variWavelength (µm)
Wavelength (µm)
able filter designs, which we will not
(b)
(a)
describe. A fore optic (i.e., telescope),
typically an unobscured form composed of two or three mirrors, is
FIGURE 3. (a) and (b) The LWIR spectra of several minerals and vegetation. All of the spectra
required to produce a high-quality
are reproduced from the Advanced Spaceborne Thermal Emission and Reflection Radiometer
image of the scene. Both afocal and
(ASTER) spectral library and the Arizona State University thermal emission spectral library
focal designs have been used. We
except for the opal spectrum, which was measured at the Jet Propulsion Laboratory's specwill not describe these in detail and
troscopy laboratory. Emissivity values are scaled and offset for clarity [2].

IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE

JUNE 2019

IEEE Geoscience and Remote Sensing Magazine - June 2019

Table of Contents for the Digital Edition of IEEE Geoscience and Remote Sensing Magazine - June 2019