336
R. Arnone et al.
by coupling these properties with a forecast circulation model. Circulation models
have advanced significantly in the last 10 years and are approaching the confidence of weather forecasting. Circulation models are driven by winds and boundary
conditions and are data assimilative of sea surface temperatures and sea surface
height.
A major forcing in coastal enviroments is the result of the physical processes such
as tides and winds and river discharge (Arnone et al., 2007). These physical processes change on scales of hours and results in advection of water masses. Although
bio-optical processes are different than physical processes, we argue that bio-optical
time scales occur on longer (on order of several days) compared to hourly time
scales of the physical forcing (Stramska et al., 1995). The coupling of the bio-optical
properties with the physical circulation models should provide a capability to forecast bio-optical properties on short times scales (days), where as at longer times
scales (several days to weeks) the bio-optical processes may be decoupled from
physical processes. For these longer time scale forecasting, other more complex
bio-optical models may be required (Jolliff et al., 2008).
Satellite ocean color products are available from several satellites such that daily
imagery is available along most coastlines. The initialization field of the coastal
environment can be reinitialized daily using updated satellite bio-optical products
such that a 24 h forecast should be possible for coastal decisons.
Our objective is to demonstrate the ability to derive bio-optical products along
coastal waters on short time scales based on coupling ocean color bio-optical products with forecast ocean circulation models. We evaluate the bio-optical forecast
using satellite imagery to determine an effective forecast probability.
19.2 Satellite Ocean Color and Circulation Models
Ocean color imagery from MODIS-Aqua was used to determine the bio-optical
properties along the coast of northern Gulf of Mexico at the Mississippi River
delta, USA for a month period in October, 2009. The Quasi Analytical Algorithm
(QAA) (Lee et al., 2002; Martinolich, 2006) was used to process the imagery into
backscattering (551) and absorption (443) and chlorophyll products. We used the
1 km products in our example, although new algorithms have been developed to
determine these properties at 250 m (Ladner et al., 2007) which show improved
capability for coastal management requiring the high resolution. The backscattering coefficient has been used to estimate an effective particle concentration, if we
assume a specific size and composition and this property can be treated as a water
mass tracer only affected by particle settling. The total absorption properties are
influenced by phytoplankton, detritus and colored dissolved organic matter. The
absorption properties are less of a conservative tracer since they are influenced by
biological processes such as growth, decay photo-oxidation, etc. However, on these
short time scales, we argue they also are minimal since diurnal changes in phytoplankton concentration can be small. This is especially appropriate in the coast
R. Arnone et al.
by coupling these properties with a forecast circulation model. Circulation models
have advanced significantly in the last 10 years and are approaching the confidence of weather forecasting. Circulation models are driven by winds and boundary
conditions and are data assimilative of sea surface temperatures and sea surface
height.
A major forcing in coastal enviroments is the result of the physical processes such
as tides and winds and river discharge (Arnone et al., 2007). These physical processes change on scales of hours and results in advection of water masses. Although
bio-optical processes are different than physical processes, we argue that bio-optical
time scales occur on longer (on order of several days) compared to hourly time
scales of the physical forcing (Stramska et al., 1995). The coupling of the bio-optical
properties with the physical circulation models should provide a capability to forecast bio-optical properties on short times scales (days), where as at longer times
scales (several days to weeks) the bio-optical processes may be decoupled from
physical processes. For these longer time scale forecasting, other more complex
bio-optical models may be required (Jolliff et al., 2008).
Satellite ocean color products are available from several satellites such that daily
imagery is available along most coastlines. The initialization field of the coastal
environment can be reinitialized daily using updated satellite bio-optical products
such that a 24 h forecast should be possible for coastal decisons.
Our objective is to demonstrate the ability to derive bio-optical products along
coastal waters on short time scales based on coupling ocean color bio-optical products with forecast ocean circulation models. We evaluate the bio-optical forecast
using satellite imagery to determine an effective forecast probability.
19.2 Satellite Ocean Color and Circulation Models
Ocean color imagery from MODIS-Aqua was used to determine the bio-optical
properties along the coast of northern Gulf of Mexico at the Mississippi River
delta, USA for a month period in October, 2009. The Quasi Analytical Algorithm
(QAA) (Lee et al., 2002; Martinolich, 2006) was used to process the imagery into
backscattering (551) and absorption (443) and chlorophyll products. We used the
1 km products in our example, although new algorithms have been developed to
determine these properties at 250 m (Ladner et al., 2007) which show improved
capability for coastal management requiring the high resolution. The backscattering coefficient has been used to estimate an effective particle concentration, if we
assume a specific size and composition and this property can be treated as a water
mass tracer only affected by particle settling. The total absorption properties are
influenced by phytoplankton, detritus and colored dissolved organic matter. The
absorption properties are less of a conservative tracer since they are influenced by
biological processes such as growth, decay photo-oxidation, etc. However, on these
short time scales, we argue they also are minimal since diurnal changes in phytoplankton concentration can be small. This is especially appropriate in the coast