IEEE Geoscience and Remote Sensing Magazine - June 2019 - 79

The type of disperser, prism, or grating plays a role in the
design space. For a classical spectrometer, there are optical
forms that can accommodate either type of dispersing element. This is also true for some imaging spectrometer designs such as the Cook reflective triplet (see [1], page 208),
but most designs have shaped the disperser to meet optical
performance goals [9]. An example of a prism-based design
is the Spatially Enhanced Broadband Array Spectrograph
System (SEBASS), first used in 1995 [10]. SEBASS uses a
prism and hyperboloid reflector in a double-pass configuration. Sodium chloride is the prism material for the LWIR
channel, with the index of refraction varying from 1.51 at
7.8 nm to 1.47 at 13.4 nm, which produces a limit of resolution that varies from 64 to 35 nm. SEBASS does not display the desired spectral-spatial uniformity because of the
wavelength-dependent dispersion, although the combination of a shaped prism and a hyperboloid mirror provides
adequate aberration control over the full spectral range
with the minimal keystone and smile mitigated by using
a curved slit. SEBASS is housed in a liquid helium-cooled
dewar to reduce self-emission and for FPA cooling. An additional drawback was the narrow scan of ! 3.6c. Table 2
summarizes several noncommercial LWIR imaging spectrometers and their optical characteristics. Additionally,
there are some commercially available imaging spectrometers such as the OWL, manufactured by Specim Spectral
Imaging; the Hyper-Cam thermal airborne hyperspectral
imager, developed by Telops; and the TASI-600, manufactured by ITRES.
The current state of the art is a modification of a concentric microlithographic projector of unit magnification
invented by J. Dyson in 1959 [11]. The form is composed
of a plano-convex thick lens and a spherical mirror, as
shown in Figure 6. Concentric design forms are readily
adaptable as imaging spectrometers because the Seidel aberrations are already corrected due to symmetry. The Dyson form is modified by replacing the concave mirror with
a concave grating and providing some relief for mounting
the slit and FPA. The addition of the grating breaks the
symmetry and introduces significant aberrations, principally astigmatism, which must be corrected in the spectrometer design.
There are two types of modified Dyson imaging spectrometers. The first is MAKO, in which a corrector lens
with a hyperboloid figure has been added to the Dyson
form near the concave diffraction grating [12]. The corrector is used for aberration control and introduces space
between the thick lens and both the slit and FPAm, facilitating mounting. MAKO is an F/1.25 system, where
the F-number (N) is 1.25. The pixels are 75 nm, with the
diffraction-limited spot size contained within a pixel at all
wavelengths. It can be operated in a wide-area scan (! 45c
scan) or in single-target modes. Another approach was
developed for the Hyperspectral Thermal Emission Spectrometer (HyTES), where the thick lens has been extensively modified to include a narrow passage for the light from
JUNE 2019

IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE

the slit and a sloped back surface that internally reflects
the returning dispersed light onto the FPA [13]. HyTES is
an F/1.6 system with an aggregated pixel of 29 nm, and
the diffraction-limited point spread function exceeds the
detector element dimensions at roughly 10 nm. Data
are acquired through whisk broom scanning with a scan
width of ! 25c. Both designs utilize single-blaze spherical
gratings manufactured using either diamond turning or
electron beam manufacturing techniques. The aberration
control of these modified Dyson sensors is excellent. Both
forms are shown in Figure 7.
We will only briefly touch on the use of interferometric
forms because they have not been used as flight instruments
to the degree that dispersive designs have. This is largely because the sensors take spectral imagery in a staring mode
with all of the spatial elements in the image cube recorded
over a relatively long period of time. The staring mode is
required because the two interferometric forms rely upon
changing the accurately known separation between two
optics to introduce a path difference that is the metric for
determining the wavelength. This is more difficult to accomplish on a moving platform as compared to a dispersive
sensor. Figure 8 illustrates both Fabry-Pérot and Michelson interferometers. An example of a Fabry-Pérot imaging
spectrometer developed for the gas detection problem is in

TABLE 2. THE INSTRUMENT CHARACTERISTICS OF THREE
DISPERSIVE IMAGING SPECTROMETERS.
SENSOR

RANGE
(nm)

U
(10 -6cm2sr)

m-SPACING
(nm)

IFOV
(mrad)

SEBASS

7.6-13.5

7.2

4.9

69-35

1.10

43.8

0.55

17.6

1.70

MAKO

7.45-13.5

1.25

HyTES

7.5-12

1.6

4.7

IFOV: instantaneous field of view.

R1
x

Concave Mirror

Plano-Convex
Lens

O
C
I

FIGURE 6. A schematic of the Dyson form composed of a plano-

convex thick lens and a spherical mirror. The radii of curvature of
each powered surface are centered on C. O is an object point, and
I is an image point. For illustration, the green line is the trace of a
single ray. The radii are related through R 2 /R 1 = n/ (n -1), where
n is the index of refraction of the lens glass.
79

IEEE Geoscience and Remote Sensing Magazine - June 2019

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