350
S. Sathyendranath and T. Platt
Science advances in fits and starts, and often takes off in directions that were
totally un-anticipated. Therefore, the authors have some latitude in thinking about
the future, and can proceed knowing that the anticipated developments might differ
from the eventual reality. The views presented in this paper are not intended to be
exhaustive; instead it is very much a personal view, moulded by personal experience.
20.2 Technological Developments
The first ocean-colour satellite, the Coastal Zone Colour Scanner (CZCS), launched
in 1978, has been succeeded since 1996 (after a gap of 10 years) by a series of satellite sensors for observing ocean colour, which have all been marked by significant
improvements to their technical specifications, compared with the CZCS (IOCCG,
1999). These include better radiometric calibration, enhanced spectral resolution,
and improved signal-to-noise ratio. All these contribute to higher radiometric precision of retrieved quantities such as water-leaving radiances, and they are all essential
to improve the precision of the products derived from the observed radiometric
properties. But of all these technological developments, it is the increased spectral
resolution with high radiometric precision that offers the potential for development
of novel products from ocean-colour data.
As is well known, ocean-colour radiometers were designed initially for the
retrieval of a single variable: chlorophyll concentration. However, it was soon recognised that visible spectral radiometers with limited spectral wavebands, such as the
CZCS, worked reasonably well only in the so-called Case 1 waters, where phytoplankton could be considered the single, independent variable responsible for
changes in optical properties of sea water, and hence, in ocean colour. Applications
to complex coastal and inland waters required additional wavebands. The time and
space scales associated with coastal processes are typically smaller than those of
open-ocean processes, and interrogating the coastal ocean at the appropriate scales
requires higher spatial and temporal resolution than for the open ocean (IOCCG,
2000).
Recognition of these limitations led to improved spectral resolution in subsequent satellite sensors, which, along with improved radiometric precision to allow
better atmospheric correction, were designed for application of ocean-colour data in
the so-called Case-2 waters – optically-complex waters often encountered in coastal
systems and fresh-water bodies – in which substances other than phytoplankton,
such as suspended sediments and yellow substances, have an important and independent influence on optical properties. Pixel resolution at ground level has also
improved over time, also contributing to improved applications in coastal waters.
Technological developments have also taken satellite sensors beyond multispectral sensors into the realm of hyper-spectral sensors: a term reserved for sensors
with spectral resolution approaching a few nanometres. Geostationary ocean-colour
sensors that allow high-frequency observations, once just a pipe dream, are now
close to becoming a reality, with the anticipated launch of the Korean satellite GOCI
in November 2009.
S. Sathyendranath and T. Platt
Science advances in fits and starts, and often takes off in directions that were
totally un-anticipated. Therefore, the authors have some latitude in thinking about
the future, and can proceed knowing that the anticipated developments might differ
from the eventual reality. The views presented in this paper are not intended to be
exhaustive; instead it is very much a personal view, moulded by personal experience.
20.2 Technological Developments
The first ocean-colour satellite, the Coastal Zone Colour Scanner (CZCS), launched
in 1978, has been succeeded since 1996 (after a gap of 10 years) by a series of satellite sensors for observing ocean colour, which have all been marked by significant
improvements to their technical specifications, compared with the CZCS (IOCCG,
1999). These include better radiometric calibration, enhanced spectral resolution,
and improved signal-to-noise ratio. All these contribute to higher radiometric precision of retrieved quantities such as water-leaving radiances, and they are all essential
to improve the precision of the products derived from the observed radiometric
properties. But of all these technological developments, it is the increased spectral
resolution with high radiometric precision that offers the potential for development
of novel products from ocean-colour data.
As is well known, ocean-colour radiometers were designed initially for the
retrieval of a single variable: chlorophyll concentration. However, it was soon recognised that visible spectral radiometers with limited spectral wavebands, such as the
CZCS, worked reasonably well only in the so-called Case 1 waters, where phytoplankton could be considered the single, independent variable responsible for
changes in optical properties of sea water, and hence, in ocean colour. Applications
to complex coastal and inland waters required additional wavebands. The time and
space scales associated with coastal processes are typically smaller than those of
open-ocean processes, and interrogating the coastal ocean at the appropriate scales
requires higher spatial and temporal resolution than for the open ocean (IOCCG,
2000).
Recognition of these limitations led to improved spectral resolution in subsequent satellite sensors, which, along with improved radiometric precision to allow
better atmospheric correction, were designed for application of ocean-colour data in
the so-called Case-2 waters – optically-complex waters often encountered in coastal
systems and fresh-water bodies – in which substances other than phytoplankton,
such as suspended sediments and yellow substances, have an important and independent influence on optical properties. Pixel resolution at ground level has also
improved over time, also contributing to improved applications in coastal waters.
Technological developments have also taken satellite sensors beyond multispectral sensors into the realm of hyper-spectral sensors: a term reserved for sensors
with spectral resolution approaching a few nanometres. Geostationary ocean-colour
sensors that allow high-frequency observations, once just a pipe dream, are now
close to becoming a reality, with the anticipated launch of the Korean satellite GOCI
in November 2009.