62
551 nm as the ratio denominator (OC3), or the ratio of R rs at either 443, 490, or 510
to R rs at 555 nm (OC4) (O’Reilly et al. 1998; Werdell et al. 2009). In both cases, the
ratio denominator is known as a hinge point since it is located at a wavelength in the
spectrum where chlorophyll absorbs little and hence reflects green light. Various
weighting mechanisms are used to select the ratio used such as choosing the highest
for any pixel, or setting cutoffs to avoid natural spectral reflectance red-shifts at
higher Chl a.
These algorithms are successfully applicable to waters where absorption or scattering by materials different from phytoplankton is minimal, but such conditions are
not always met in coastal waters where suspended material and colored dissolved
organic matter (CDOM) can play a large part in determining the effective R rs spectrum. The exponential decay spectrum of CDOM, absorbing strongly in the blue but
minimally in the red, can lead these simple algorithms astray in areas of high CDOM
content. Since CDOM absorbs most strongly in the blue, large discrepancies
between satellite-retrieved Chl a values and in situ measurements result when estimating Chl a using algorithms dependent on blue to green band ratios (Del Castillo
2005; Müller-Karger et al. 2005).
Spectral band-difference algorithms can allow accurate visualization of dense
phytoplankton blooms. These differ from the band-ratio family in that baselines
extending from the low reflectance PAR region across the high reflectance near
infrared (NIR) into the short wave infrared (SWIR) allow direct estimation of the
NIR peak of interest. Such approaches have been exploited successfully for detection of various algal blooms of practical concern. The fluorescence line height
(FLH) algorithm (667, 678, and 746 nm) exploits in vivo fluorescence at 678 nm
while the floating algae index (FAI), operating at 667, 859, and 1240 nm, and its
modifications exploit the so-called red-edge across the red-NIR spectrum (BlondeauPattissier et al. 2014).
Extensive mats of floating Sargassum sp. appearing of late in the Caribbean have
become a nuisance to nearshore marine operations as these rafts become entrained
along beaches and harbors. Mats extending out tens of meters from the coast along
windward shorelines particularly hamper small boat operations impacting tourism
and fishing. Sargassum blooms have been successfully tracked and operational
products are available at scales relevant to stakeholders: (http://www.caricoos.org/
oceans/observation/modis_aqua/ECARIBE/afai Accessed 8/25/2017). Detection
depends on a floating algae index across the vegetation red edge using the baseline
between 667 and 1240 nm (or 1640) and the diagnostic peak at 859 nm where floating algal biomass is highly reflective (https://eos.org/features/sargassum-watchwarns-of-incoming-seaweed).
Specific products for mapping harmful algal blooms via satellite imagery use the
MODIS and VIIRS chlorophyll data products to inform statistical models augmented by periodic sample collection as discussed below in Chap. 6.
2 Electronic Sensors and Instruments for Coastal Ocean Observing
551 nm as the ratio denominator (OC3), or the ratio of R rs at either 443, 490, or 510
to R rs at 555 nm (OC4) (O’Reilly et al. 1998; Werdell et al. 2009). In both cases, the
ratio denominator is known as a hinge point since it is located at a wavelength in the
spectrum where chlorophyll absorbs little and hence reflects green light. Various
weighting mechanisms are used to select the ratio used such as choosing the highest
for any pixel, or setting cutoffs to avoid natural spectral reflectance red-shifts at
higher Chl a.
These algorithms are successfully applicable to waters where absorption or scattering by materials different from phytoplankton is minimal, but such conditions are
not always met in coastal waters where suspended material and colored dissolved
organic matter (CDOM) can play a large part in determining the effective R rs spectrum. The exponential decay spectrum of CDOM, absorbing strongly in the blue but
minimally in the red, can lead these simple algorithms astray in areas of high CDOM
content. Since CDOM absorbs most strongly in the blue, large discrepancies
between satellite-retrieved Chl a values and in situ measurements result when estimating Chl a using algorithms dependent on blue to green band ratios (Del Castillo
2005; Müller-Karger et al. 2005).
Spectral band-difference algorithms can allow accurate visualization of dense
phytoplankton blooms. These differ from the band-ratio family in that baselines
extending from the low reflectance PAR region across the high reflectance near
infrared (NIR) into the short wave infrared (SWIR) allow direct estimation of the
NIR peak of interest. Such approaches have been exploited successfully for detection of various algal blooms of practical concern. The fluorescence line height
(FLH) algorithm (667, 678, and 746 nm) exploits in vivo fluorescence at 678 nm
while the floating algae index (FAI), operating at 667, 859, and 1240 nm, and its
modifications exploit the so-called red-edge across the red-NIR spectrum (BlondeauPattissier et al. 2014).
Extensive mats of floating Sargassum sp. appearing of late in the Caribbean have
become a nuisance to nearshore marine operations as these rafts become entrained
along beaches and harbors. Mats extending out tens of meters from the coast along
windward shorelines particularly hamper small boat operations impacting tourism
and fishing. Sargassum blooms have been successfully tracked and operational
products are available at scales relevant to stakeholders: (http://www.caricoos.org/
oceans/observation/modis_aqua/ECARIBE/afai Accessed 8/25/2017). Detection
depends on a floating algae index across the vegetation red edge using the baseline
between 667 and 1240 nm (or 1640) and the diagnostic peak at 859 nm where floating algal biomass is highly reflective (https://eos.org/features/sargassum-watchwarns-of-incoming-seaweed).
Specific products for mapping harmful algal blooms via satellite imagery use the
MODIS and VIIRS chlorophyll data products to inform statistical models augmented by periodic sample collection as discussed below in Chap. 6.
2 Electronic Sensors and Instruments for Coastal Ocean Observing