The Availability of Timely Global Oceanographic Data from Satellites
9
to the IOCCG. Although both sensors and algorithms for processing their output are
currently evolving with great rapidity, the evaluation of data from optically complex
seawater ideally requires both special algorithms and also sensors able to distinguish the
relative contributions to an optical property of suspended material, of CDOM, and of
phytoplankton.
Although SeaWiFS images are still being obtained and made available, global chlorophyll images are also obtained routinely from the MODIS (Moderate-resolution Imaging
Spectroradiometer) sensors that are operated by NASA’s Earth Observing System for the
U.S. Global Change Program. These are carried aboard the NASA Terra and Aqua satellites: these began work in tandem so that data from every area of the planet were returned
twice daily—during local morning and afternoon periods—but, because of technical
problems, we now obtain these data only from the Aqua satellite.
After the application of cloud and atmospheric correction algorithms, these data
deliver 36 wavelength-specific parameters that describe water-leaving irradiance and total
absorption coefficients; these comprise computations of chlorophyll-a to first optic depth,
fluorescence efficiency, CDOM, and calcite concentrations. Otherwise, four parameters
are used to compute SST and eight for the computation of primary production while
yet other channels are used for data quality control and environmental conditions: sea
surface temperature, opacity of cloud cover, aerosol properties, and photosynthetically
available radiation.
Unlike SeaWiFS, MODIS measures chlorophyll not only by absorption, but also by
reference to fluorescence (F), the emitted light energy that is not used for photosynthesis.
Estimates of F will be especially useful in estimation of chlorophyll where other materials
that scatter and absorb light, such as sediments and CDOM, are abundant. Finally,
progress has already been made in deploying two algorithms for computation of primary
productivity from MODIS data: parameter P1 is based on the vertically generalized model
of Behrenfeld and Falkowski (1997b), integrated over the photic zone, and parameter
P2 on a Howard, Yoder, and Ryan version of the Epply/Peterson polynomial, integrated
over the mixed layer depth. You can obtain further details of the processing and output
of MODIS data from the MODIS Users Guide available on the Web.
At first order, we may assume that the indicated chlorophyll values from SeaWiFS or
MODIS represent mixed-layer pigments and that subsurface chlorophyll maxima deeper
than ∼25 m are not detected directly. For this reason, if uplift and illumination of the
nutricline lead to a subsurface bloom, this may not be observed in water-leaving radiation;
however, we can expect that such a bloom will enhance mixed-layer “background”
chlorophyll values by the same vertical mixing process that maintains the mixed layer
itself. This enhanced background may then be regionally detectable in the chlorophyll
field, though this may not represent the true intensity of the subsurface bloom. However,
as will be discussed later, we do possess a sufficient archive of observations of chlorophyll
profiles to make a reasonable prediction of the seasonally variable subsurface chlorophyll
field in all parts of the oceans.
It is difficult now to remember how ignorant we were of the extent, variability, and
seasonality of algal blooms prior to their global visualization by the CZCS (Banse and
McClain, 1986; Brock et al., 1993; Muller-Karger et al., 1989; Longhurst, 1993). CZCS data
were at first available simply as a brochure of a few striking global and regional images,
but the files soon became available as a climatology of monthly means for 1978–1986 on
a 1
grid covering all oceans and seas, comprising 42,732 data points, of which many were
blank because sensors were often switched off and because of cloud cover. This was the
archive used by Banse and English (1994) in their groundbreaking study of the seasonality
of near-surface chlorophyll in the ocean, but we are now immeasurably richer, having
already access on the NASA Goddard site to 8 full years of global SeaWiFS and 4 years
of MODIS images, including SST. These data have all been used to compute primary
9
to the IOCCG. Although both sensors and algorithms for processing their output are
currently evolving with great rapidity, the evaluation of data from optically complex
seawater ideally requires both special algorithms and also sensors able to distinguish the
relative contributions to an optical property of suspended material, of CDOM, and of
phytoplankton.
Although SeaWiFS images are still being obtained and made available, global chlorophyll images are also obtained routinely from the MODIS (Moderate-resolution Imaging
Spectroradiometer) sensors that are operated by NASA’s Earth Observing System for the
U.S. Global Change Program. These are carried aboard the NASA Terra and Aqua satellites: these began work in tandem so that data from every area of the planet were returned
twice daily—during local morning and afternoon periods—but, because of technical
problems, we now obtain these data only from the Aqua satellite.
After the application of cloud and atmospheric correction algorithms, these data
deliver 36 wavelength-specific parameters that describe water-leaving irradiance and total
absorption coefficients; these comprise computations of chlorophyll-a to first optic depth,
fluorescence efficiency, CDOM, and calcite concentrations. Otherwise, four parameters
are used to compute SST and eight for the computation of primary production while
yet other channels are used for data quality control and environmental conditions: sea
surface temperature, opacity of cloud cover, aerosol properties, and photosynthetically
available radiation.
Unlike SeaWiFS, MODIS measures chlorophyll not only by absorption, but also by
reference to fluorescence (F), the emitted light energy that is not used for photosynthesis.
Estimates of F will be especially useful in estimation of chlorophyll where other materials
that scatter and absorb light, such as sediments and CDOM, are abundant. Finally,
progress has already been made in deploying two algorithms for computation of primary
productivity from MODIS data: parameter P1 is based on the vertically generalized model
of Behrenfeld and Falkowski (1997b), integrated over the photic zone, and parameter
P2 on a Howard, Yoder, and Ryan version of the Epply/Peterson polynomial, integrated
over the mixed layer depth. You can obtain further details of the processing and output
of MODIS data from the MODIS Users Guide available on the Web.
At first order, we may assume that the indicated chlorophyll values from SeaWiFS or
MODIS represent mixed-layer pigments and that subsurface chlorophyll maxima deeper
than ∼25 m are not detected directly. For this reason, if uplift and illumination of the
nutricline lead to a subsurface bloom, this may not be observed in water-leaving radiation;
however, we can expect that such a bloom will enhance mixed-layer “background”
chlorophyll values by the same vertical mixing process that maintains the mixed layer
itself. This enhanced background may then be regionally detectable in the chlorophyll
field, though this may not represent the true intensity of the subsurface bloom. However,
as will be discussed later, we do possess a sufficient archive of observations of chlorophyll
profiles to make a reasonable prediction of the seasonally variable subsurface chlorophyll
field in all parts of the oceans.
It is difficult now to remember how ignorant we were of the extent, variability, and
seasonality of algal blooms prior to their global visualization by the CZCS (Banse and
McClain, 1986; Brock et al., 1993; Muller-Karger et al., 1989; Longhurst, 1993). CZCS data
were at first available simply as a brochure of a few striking global and regional images,
but the files soon became available as a climatology of monthly means for 1978–1986 on
a 1
grid covering all oceans and seas, comprising 42,732 data points, of which many were
blank because sensors were often switched off and because of cloud cover. This was the
archive used by Banse and English (1994) in their groundbreaking study of the seasonality
of near-surface chlorophyll in the ocean, but we are now immeasurably richer, having
already access on the NASA Goddard site to 8 full years of global SeaWiFS and 4 years
of MODIS images, including SST. These data have all been used to compute primary