SATELLITE MEASUREMENTS
167
sensors. However, as indicated in Figure 7 a visible waveband radiometer
will “see” only down to the level where the irradiance is about 1/3 of its
surface value. The satellite measurements are unlikely to record accurately,
if at all, any phytoplankton below this level, such as those contributing to the
deep chlorophyll maximum found at the base of a mixed layer after nutrients
have been used up from the mixed layer. Similarly if a second, low-light,
species develops below the main bloom the satellite will not be able to detect
them.
There is undoubtedly a large amount of valuable information for ocean
models to be found in the ocean colour data products from satellites.
However considerably more research is needed to learn how best to inject
that information into the models. A particularly enticing prize is to combine
the satellite measurements with ocean carbon cycle models to be able to
estimate with some confidence the rates of primary production occurring in
the sea. Allied to this is the potential for improving our knowledge of how
pCO 2 (representing the amount of CO 2 dissolved in the surface water) is
distributed, leading to better estimates of air-sea fluxes of CO 2 . Finally we
should not overlook the much simpler application of using ocean colour as a
tracer of mesoscale eddies. There is a need to develop techniques to
assimilate this information so that eddy-resolving ocean circulation models
are guided to present eddies in the right place at the right time.
3.3
Sea surface temperature
3.3.1
Diverse methods for measuring sea surface temperature
Sea surface temperature (SST) can be measured in a variety of ways,
using sensors on both satellites and in situ platforms (Robinson & Donlon,
2003). Sampling from in situ platforms can generally be performed at high
frequency whereas most satellite methods are severely restricted by orbit
constraints to long sampling intervals of several hours or more. On the other
hand remote sensors are capable of wide synoptic spatial coverage at fine
spatial detail down to 1 km resolution when unobstructed by clouds, while
all the in situ methods sample very sparsely, and may miss some regions
altogether. Table 3 lists the different classes of satellite-based methods and
the typical absolute accuracy of measurements which they can achieve.
Relative accuracy (that is the smallest temperature difference that can be
detected confidently within a given image from a single overpass) may be
somewhat better than the absolute value quoted.
167
sensors. However, as indicated in Figure 7 a visible waveband radiometer
will “see” only down to the level where the irradiance is about 1/3 of its
surface value. The satellite measurements are unlikely to record accurately,
if at all, any phytoplankton below this level, such as those contributing to the
deep chlorophyll maximum found at the base of a mixed layer after nutrients
have been used up from the mixed layer. Similarly if a second, low-light,
species develops below the main bloom the satellite will not be able to detect
them.
There is undoubtedly a large amount of valuable information for ocean
models to be found in the ocean colour data products from satellites.
However considerably more research is needed to learn how best to inject
that information into the models. A particularly enticing prize is to combine
the satellite measurements with ocean carbon cycle models to be able to
estimate with some confidence the rates of primary production occurring in
the sea. Allied to this is the potential for improving our knowledge of how
pCO 2 (representing the amount of CO 2 dissolved in the surface water) is
distributed, leading to better estimates of air-sea fluxes of CO 2 . Finally we
should not overlook the much simpler application of using ocean colour as a
tracer of mesoscale eddies. There is a need to develop techniques to
assimilate this information so that eddy-resolving ocean circulation models
are guided to present eddies in the right place at the right time.
3.3
Sea surface temperature
3.3.1
Diverse methods for measuring sea surface temperature
Sea surface temperature (SST) can be measured in a variety of ways,
using sensors on both satellites and in situ platforms (Robinson & Donlon,
2003). Sampling from in situ platforms can generally be performed at high
frequency whereas most satellite methods are severely restricted by orbit
constraints to long sampling intervals of several hours or more. On the other
hand remote sensors are capable of wide synoptic spatial coverage at fine
spatial detail down to 1 km resolution when unobstructed by clouds, while
all the in situ methods sample very sparsely, and may miss some regions
altogether. Table 3 lists the different classes of satellite-based methods and
the typical absolute accuracy of measurements which they can achieve.
Relative accuracy (that is the smallest temperature difference that can be
detected confidently within a given image from a single overpass) may be
somewhat better than the absolute value quoted.