Hyperspectral Sensors and Applications
23
of HyMap, state that SNR values are above 500:1 at 2200 nm, and 1000:1 in
the VNIR region. For the GER sensor, values of 5000:1 for the visible and
NIR regions and 500: 1 for the SWIR region, have been reported (Mackin and
Munday 1988). The SNR also varies depending upon the nature of the surface.
In a study of soils in Israel (Ben-Dor and Levin 2000), the SNR for sensors
varied from 20:1 to 120:1 for light targets and 1:1 to 5:1 for dark targets.
The development and use of airborne hyperspectral sensors has continued
as they provide flexibility in the acquisition of data, particularly in terms of
temporal frequency and spatial coverage. Specifically, data can be acquired
when conditions (e. g., local weather) are optimal, and parameters (e. g., the
area of coverage, waveband configurations and spatial resolutions) can be determined prior to acquisition. Even so, the expense of acquiring data from
airborne sensors still limits the construction of multi-temporal datasets and
generally reduces coverage to relatively restricted areas. Many studies using
such data have therefore been developed for specific and often narrow applications. A further disadvantage of using airborne sensors has been that aircraft
motion and sun-surface-sensor geometries impact negatively on the quality
of the data acquired. For these reasons, there has been a drive to advance
spaceborne missions with hyperspectral capability.
1.3.2
Spaceborne sensors
Although airborne hyperspectral sensors have undoubtedly made a valuable
contribution to Earth observation research and monitoring and satisfied many
of the demands of the user community, the arguments for spaceborne observations have been compelling. Firstly, such sensors allow regular, repeated
observations and coverage over wider areas. Secondly, variations in both atmospheric and bi-directional effects across the scans, and distortions resulting
from sensor motion, are also reduced.
The drive towards spaceborne hyperspectral data was hampered initially
as a planned Space Shuttle imaging spectrometer was cancelled following the
Challenger accident. Delays also occurred in implementing the Earth Observing System (EOS) High Resolution Imaging Spectrometer (HRIS) (Goetz and
Herring 1989) due to EOS budget constraints and the failure of the LEWIS satellite to achieve orbit. The first satellite-based imaging spectrometer available for
general usage was therefore the Moderate Resolution Imaging Spectrometer
(MODIS) on both the EOS TERRA-l and AQUA satellites, which were launched
in 1999 and 2002 respectively. The TERRA-l Advanced Spaceborne Thermal
Emission and Reflection (ASTER) sensor and the ENVISAT Medium Resolution Imaging Spectrometer (MERIS) sensor (launched 2002) also provide
hyperspectral capability, albeit in selected wavelength regions only. Both the
MODIS and MERIS sensors provide near global coverage at coarse (> 250 m)
spatial resolution.
Hyperion was the first hyperspectral sensor providing data across the full
reflectance spectral region and was launched in November, 2000, onboard
the NASA Earth Observing-l (EO-1) platform. In October, 2001, the Compact
23
of HyMap, state that SNR values are above 500:1 at 2200 nm, and 1000:1 in
the VNIR region. For the GER sensor, values of 5000:1 for the visible and
NIR regions and 500: 1 for the SWIR region, have been reported (Mackin and
Munday 1988). The SNR also varies depending upon the nature of the surface.
In a study of soils in Israel (Ben-Dor and Levin 2000), the SNR for sensors
varied from 20:1 to 120:1 for light targets and 1:1 to 5:1 for dark targets.
The development and use of airborne hyperspectral sensors has continued
as they provide flexibility in the acquisition of data, particularly in terms of
temporal frequency and spatial coverage. Specifically, data can be acquired
when conditions (e. g., local weather) are optimal, and parameters (e. g., the
area of coverage, waveband configurations and spatial resolutions) can be determined prior to acquisition. Even so, the expense of acquiring data from
airborne sensors still limits the construction of multi-temporal datasets and
generally reduces coverage to relatively restricted areas. Many studies using
such data have therefore been developed for specific and often narrow applications. A further disadvantage of using airborne sensors has been that aircraft
motion and sun-surface-sensor geometries impact negatively on the quality
of the data acquired. For these reasons, there has been a drive to advance
spaceborne missions with hyperspectral capability.
1.3.2
Spaceborne sensors
Although airborne hyperspectral sensors have undoubtedly made a valuable
contribution to Earth observation research and monitoring and satisfied many
of the demands of the user community, the arguments for spaceborne observations have been compelling. Firstly, such sensors allow regular, repeated
observations and coverage over wider areas. Secondly, variations in both atmospheric and bi-directional effects across the scans, and distortions resulting
from sensor motion, are also reduced.
The drive towards spaceborne hyperspectral data was hampered initially
as a planned Space Shuttle imaging spectrometer was cancelled following the
Challenger accident. Delays also occurred in implementing the Earth Observing System (EOS) High Resolution Imaging Spectrometer (HRIS) (Goetz and
Herring 1989) due to EOS budget constraints and the failure of the LEWIS satellite to achieve orbit. The first satellite-based imaging spectrometer available for
general usage was therefore the Moderate Resolution Imaging Spectrometer
(MODIS) on both the EOS TERRA-l and AQUA satellites, which were launched
in 1999 and 2002 respectively. The TERRA-l Advanced Spaceborne Thermal
Emission and Reflection (ASTER) sensor and the ENVISAT Medium Resolution Imaging Spectrometer (MERIS) sensor (launched 2002) also provide
hyperspectral capability, albeit in selected wavelength regions only. Both the
MODIS and MERIS sensors provide near global coverage at coarse (> 250 m)
spatial resolution.
Hyperion was the first hyperspectral sensor providing data across the full
reflectance spectral region and was launched in November, 2000, onboard
the NASA Earth Observing-l (EO-1) platform. In October, 2001, the Compact