24
1: Richard lucas, Aled Rowlands, Olaf Niemann, Ray Merton
High Resolution Imaging Spectrometer (CHRIS) was also deployed onboard
the Project for On-Board Autonomy (PROBA) small satellite mission. CHRIS
was designed for a broad base of applications providing hyperspectral data
within the 410-lOS0 nm wavelength regions and over a wide range of spectral
and spatial configurations, thereby providing hyperspectral Bidirectional Reflectance Distribution Function (BRDF) observations of the Earth's surface at
-SSo, -36°, 0° (nadir), +36°, +SSo along track look angles. Orbview-4 was to
provide 1 m panchromatic and 4 m multispectral data as well as a 200 channel
hyperspectral sensor, although the sensor was lost in a launch failure. A number
of satellite systems also remain in the development phase (Table 1.3), including
ESA's Visible and Infrared Thermal Imaging Spectrometer (VIRTIS-H) and
Land Surface Processes and Interactions Mission (LSPIM) (Buckingham et
al. 2002), and SPECTRA, whilst research is underway into the development
of ultraspectral sensors which will provide data in thousands of very narrow
spectral bands giving sufficient spectral sensitivity for identifying specific constituents (e.g. aerosols and gas plumes). The Canadian Space Agency (CSA)
along with the Canadian Centre for Remote Sensing (CCRS) and industry is
also developing a targeted hyperspectral imager for the International Space
Station (ISS) for launch on a small satellite (Buckingham et al. 2002).
The benefits of using data from spaceborne hyperspectral sensors are only
just being realised with the recent launch of satellite-based sensors. The Hyperion mission has provided the first opportunity to explore the potential of
multi-temporal hyperspectral data for Earth surface observation and to test
synergy with other satellite systems (e.g., LANDSAT-7 ETM+). Even so, this
sensor remains more an experimental research tool rather than one that is
ready for general applications development, due largely to the limited swath
width and the generally poorly understood radiometry. The CHRIS sensor,
although not observing in SWIR regions, has given a greater opportunity to
explore the bidirectional capabilities and information content of hyper spectral
data with applications in land surface biochemical and biophysical property
retrieval and water characterisation.
At more synoptic spatial scales, the coarser spatial resolution sensors (e. g.,
MODIS) with hyperspectral capability provide a greater opportunity for multitemporal observations globally. A particular advantage of these sensors is that
they provide consistent observations over an extended time period, relative
consistency and/or knowledge of viewing angles, and uniformity in illumination and image geometry (van der Meer 2001). The following section provides
an overview of the characteristics of these sensors that renders them unique
in the context of Earth observation.
1.3.2.1
Spectral Regions, Resolutions and Bidirectional Capability
Although not providing a continuous spectral profile, the MODIS and ASTER
image the spectral region from the visible through to the SWIR and TIR regions
and at variable sampling intervals. The MERIS, although observing in only the
390-1040 nm region and at band widths of2.S-12.S nm, is fully programmable
1: Richard lucas, Aled Rowlands, Olaf Niemann, Ray Merton
High Resolution Imaging Spectrometer (CHRIS) was also deployed onboard
the Project for On-Board Autonomy (PROBA) small satellite mission. CHRIS
was designed for a broad base of applications providing hyperspectral data
within the 410-lOS0 nm wavelength regions and over a wide range of spectral
and spatial configurations, thereby providing hyperspectral Bidirectional Reflectance Distribution Function (BRDF) observations of the Earth's surface at
-SSo, -36°, 0° (nadir), +36°, +SSo along track look angles. Orbview-4 was to
provide 1 m panchromatic and 4 m multispectral data as well as a 200 channel
hyperspectral sensor, although the sensor was lost in a launch failure. A number
of satellite systems also remain in the development phase (Table 1.3), including
ESA's Visible and Infrared Thermal Imaging Spectrometer (VIRTIS-H) and
Land Surface Processes and Interactions Mission (LSPIM) (Buckingham et
al. 2002), and SPECTRA, whilst research is underway into the development
of ultraspectral sensors which will provide data in thousands of very narrow
spectral bands giving sufficient spectral sensitivity for identifying specific constituents (e.g. aerosols and gas plumes). The Canadian Space Agency (CSA)
along with the Canadian Centre for Remote Sensing (CCRS) and industry is
also developing a targeted hyperspectral imager for the International Space
Station (ISS) for launch on a small satellite (Buckingham et al. 2002).
The benefits of using data from spaceborne hyperspectral sensors are only
just being realised with the recent launch of satellite-based sensors. The Hyperion mission has provided the first opportunity to explore the potential of
multi-temporal hyperspectral data for Earth surface observation and to test
synergy with other satellite systems (e.g., LANDSAT-7 ETM+). Even so, this
sensor remains more an experimental research tool rather than one that is
ready for general applications development, due largely to the limited swath
width and the generally poorly understood radiometry. The CHRIS sensor,
although not observing in SWIR regions, has given a greater opportunity to
explore the bidirectional capabilities and information content of hyper spectral
data with applications in land surface biochemical and biophysical property
retrieval and water characterisation.
At more synoptic spatial scales, the coarser spatial resolution sensors (e. g.,
MODIS) with hyperspectral capability provide a greater opportunity for multitemporal observations globally. A particular advantage of these sensors is that
they provide consistent observations over an extended time period, relative
consistency and/or knowledge of viewing angles, and uniformity in illumination and image geometry (van der Meer 2001). The following section provides
an overview of the characteristics of these sensors that renders them unique
in the context of Earth observation.
1.3.2.1
Spectral Regions, Resolutions and Bidirectional Capability
Although not providing a continuous spectral profile, the MODIS and ASTER
image the spectral region from the visible through to the SWIR and TIR regions
and at variable sampling intervals. The MERIS, although observing in only the
390-1040 nm region and at band widths of2.S-12.S nm, is fully programmable