2 Satellite Water Colour Observations in African Seas
35
(Gower et al. 2006; Gower and King 2011) and “superblooms” of Antarctic diatoms
associated with platelet ice (Gower and King 2007b).
The MCI is computed as radiance at 709 nm above a linear baseline defined by
radiances at 681 and 753 nm (Gower et al. 2005, 2008b). As for FLH, this is calculated
only for pixels for which radiance at 865 nm is less than 15 mW m
− 2 sr
−1 nm
−1 to
eliminate land pixels and areas of strong sun glint, haze or cloud:
if L 865 < 15 mW m
−2 sr
−1 nm
−1 ,
then MCI = L 709 − L 681 − 0.389 (L 753 − L 681 )
(2.3)
where L 865 represents Level 1 radiances (as measured at the satellite) at a wavelength
of 865 nm, and similarly for other wavelengths, and the factor 0.389 represents the
wavelength ratio (709–681)/(753–681).
The MCI therefore indicates an excess radiance at 709 nm above this baseline,
which often indicates a water-leaving radiance spectrum with a peak at 709 nm. Models indicate that this type of spectrum is characteristic of intense surface plankton
blooms in which high concentrations of phytoplankton are distributed in near-surface
waters. In this case, the absorption by chlorophyll reduces radiance at wavelengths
shorter than 700 nm, while absorption by water reduces radiance at wavelengths
longer than 720 nm, leading to a radiance peak at the wavelength of minimum absorption, near 709 nm. Models also show that vegetation under a shallow layer of
water, including coral reefs, can give rise to a spectral peak near 709 nm, also giving
a positive MCI signal.
MCI can also be high due to presence of a “red-edge” in the spectrum. At this
“edge,” water-leaving radiance shows a step increase near 700 nm with increasing
wavelength. This type of spectrum is characteristic of land vegetation, for which the
red-edge (maximum rate of increase of observed radiance with wavelength) usually
occurs at a longer wavelength of about 720 nm. Over water with MERIS, we observe
an apparent step increase between 681 and 709 nm. The MCI as defined in Eq. 2.1
will give a high value in this case as well as in the case of a peak. We interpret
red-edge type spectra as showing presence of buoyant slicks of either phytoplankton
or macroalgae such as pelagic Sargassum.
We make use of daily, global composites of MCI data produced by the Grid
Processing on Demand (G-POD) facility of ESA (Gower et al. 2008b). These have
a spatial resolution of 5 km. Each composite pixel shows the maximum MCI value
computed for any RR MERIS (1.2 km) pixel assigned to that composite pixel on
that day. The reduced spatial resolution results in smaller blooms being spatially
distorted, but the use of the maximum values in the composite preserves the record
of their occurrence.
Daily composites are combined into monthly products, also at 5 km resolution,
recording the maximum MCI value measured in the month. This gives complete
spatial coverage in most areas and preserves information on detected blooms and
vegetation. The monthly products also show presence or absence of water, and so
can be used to measure water area.
The monthly composites can also be analyzed for “total MERIS count” by computing the sum of the number of MCI values above a threshold, multiplied by the
35
(Gower et al. 2006; Gower and King 2011) and “superblooms” of Antarctic diatoms
associated with platelet ice (Gower and King 2007b).
The MCI is computed as radiance at 709 nm above a linear baseline defined by
radiances at 681 and 753 nm (Gower et al. 2005, 2008b). As for FLH, this is calculated
only for pixels for which radiance at 865 nm is less than 15 mW m
− 2 sr
−1 nm
−1 to
eliminate land pixels and areas of strong sun glint, haze or cloud:
if L 865 < 15 mW m
−2 sr
−1 nm
−1 ,
then MCI = L 709 − L 681 − 0.389 (L 753 − L 681 )
(2.3)
where L 865 represents Level 1 radiances (as measured at the satellite) at a wavelength
of 865 nm, and similarly for other wavelengths, and the factor 0.389 represents the
wavelength ratio (709–681)/(753–681).
The MCI therefore indicates an excess radiance at 709 nm above this baseline,
which often indicates a water-leaving radiance spectrum with a peak at 709 nm. Models indicate that this type of spectrum is characteristic of intense surface plankton
blooms in which high concentrations of phytoplankton are distributed in near-surface
waters. In this case, the absorption by chlorophyll reduces radiance at wavelengths
shorter than 700 nm, while absorption by water reduces radiance at wavelengths
longer than 720 nm, leading to a radiance peak at the wavelength of minimum absorption, near 709 nm. Models also show that vegetation under a shallow layer of
water, including coral reefs, can give rise to a spectral peak near 709 nm, also giving
a positive MCI signal.
MCI can also be high due to presence of a “red-edge” in the spectrum. At this
“edge,” water-leaving radiance shows a step increase near 700 nm with increasing
wavelength. This type of spectrum is characteristic of land vegetation, for which the
red-edge (maximum rate of increase of observed radiance with wavelength) usually
occurs at a longer wavelength of about 720 nm. Over water with MERIS, we observe
an apparent step increase between 681 and 709 nm. The MCI as defined in Eq. 2.1
will give a high value in this case as well as in the case of a peak. We interpret
red-edge type spectra as showing presence of buoyant slicks of either phytoplankton
or macroalgae such as pelagic Sargassum.
We make use of daily, global composites of MCI data produced by the Grid
Processing on Demand (G-POD) facility of ESA (Gower et al. 2008b). These have
a spatial resolution of 5 km. Each composite pixel shows the maximum MCI value
computed for any RR MERIS (1.2 km) pixel assigned to that composite pixel on
that day. The reduced spatial resolution results in smaller blooms being spatially
distorted, but the use of the maximum values in the composite preserves the record
of their occurrence.
Daily composites are combined into monthly products, also at 5 km resolution,
recording the maximum MCI value measured in the month. This gives complete
spatial coverage in most areas and preserves information on detected blooms and
vegetation. The monthly products also show presence or absence of water, and so
can be used to measure water area.
The monthly composites can also be analyzed for “total MERIS count” by computing the sum of the number of MCI values above a threshold, multiplied by the