556
peak in river flow. By tracking the dye with continuous underway sampling
using towed vehicles, biological and chemical transformations can be
observed in a Lagrangian perspective. Tracking the dye makes it possible to
distinguish biogeochemical processes from physical processes that transport
material in the buoyant plume, or mixing that merely dilutes it. Furthermore,
the physical structure of river plumes differs in upwelling and downwelling
wind conditions. Downwelling leads to a narrow near-shore buoyant plume
that is thick, typically bottom attached, and rapidly transports material
alongshore. Under these conditions enhanced turbulence retains particulate
matter throughout the water column causing low light levels that constrain
phytoplankton growth and colored DOM photo-bleaching rates. During
upwelling, plumes can detach from the coast, spread, thin, and transport
water directly offshore (Munchow and Garvine 1993, Fong and Geyer 2001,
Hallock and Marmorino 2002). The enhanced stratification reduces
turbulence and particulate matter settles out of the plume, elevating light
levels and fostering phytoplankton growth, photo-bleaching, and the
potential for bio-accumulation of metals.
The motivations for making ocean modeling an integral component of
LATTE are rather different to those for CBLAST. The principal observing
systems used in CBLAST (moorings and MVCO/ASIT) were fixed
platforms whose locations had to be chosen in advance of the experiment. In
LATTE, the use of relocatable sampling methods (dye, AUVs, towed
bodies) enables the observing system to be adapted to the flow, and ocean
forecasting becomes of rather greater utility.
The potential value of forecasting the plume trajectory under varying
wind conditions became evident in the LATTE pilot experiment in April
2004. The first dye release coincided with the onset of upwelling winds that
drove the patch onto the south shore of Long Island, prematurely curtailing
the Lagrangian experiment. A second dye injection was made shortly
thereafter, and a switch to downwelling winds favored the formation of a
more classical coastally trapped plume that carried the dye south along the
Jersey Shore.
The LATTE program is in its infancy having passed only its first
milestone, namely the 2004 pilot program dye release. Accordingly, only
preliminary modeling results for LATTE are available. These will be
presented in section 3 as an illustration of some of the issues to be addressed
for coastal ocean forecasting in this situation. The data synthesis aspects of
the modeling, not yet realized, are quite different from CBLAST. Firstly,
variational data assimilation (Moore et al. 2004, Weaver et al. 2003) applied
to the measured dye distribution will be used to provide high a resolution
hindcast of the evolving plume trajectory and physical structure. Secondly, a
coupled physical/biological model that explicitly computes the bio-optical
features of suspended and dissolved matter (Bissett et al. 1999a, 1999b) will
be used to hindcast the depth-dependent distribution of phytoplankton and
DOM and predict the inherent optical properties of the water column. This
JOHN L. WILKIN AND LYON LANEROLLE
peak in river flow. By tracking the dye with continuous underway sampling
using towed vehicles, biological and chemical transformations can be
observed in a Lagrangian perspective. Tracking the dye makes it possible to
distinguish biogeochemical processes from physical processes that transport
material in the buoyant plume, or mixing that merely dilutes it. Furthermore,
the physical structure of river plumes differs in upwelling and downwelling
wind conditions. Downwelling leads to a narrow near-shore buoyant plume
that is thick, typically bottom attached, and rapidly transports material
alongshore. Under these conditions enhanced turbulence retains particulate
matter throughout the water column causing low light levels that constrain
phytoplankton growth and colored DOM photo-bleaching rates. During
upwelling, plumes can detach from the coast, spread, thin, and transport
water directly offshore (Munchow and Garvine 1993, Fong and Geyer 2001,
Hallock and Marmorino 2002). The enhanced stratification reduces
turbulence and particulate matter settles out of the plume, elevating light
levels and fostering phytoplankton growth, photo-bleaching, and the
potential for bio-accumulation of metals.
The motivations for making ocean modeling an integral component of
LATTE are rather different to those for CBLAST. The principal observing
systems used in CBLAST (moorings and MVCO/ASIT) were fixed
platforms whose locations had to be chosen in advance of the experiment. In
LATTE, the use of relocatable sampling methods (dye, AUVs, towed
bodies) enables the observing system to be adapted to the flow, and ocean
forecasting becomes of rather greater utility.
The potential value of forecasting the plume trajectory under varying
wind conditions became evident in the LATTE pilot experiment in April
2004. The first dye release coincided with the onset of upwelling winds that
drove the patch onto the south shore of Long Island, prematurely curtailing
the Lagrangian experiment. A second dye injection was made shortly
thereafter, and a switch to downwelling winds favored the formation of a
more classical coastally trapped plume that carried the dye south along the
Jersey Shore.
The LATTE program is in its infancy having passed only its first
milestone, namely the 2004 pilot program dye release. Accordingly, only
preliminary modeling results for LATTE are available. These will be
presented in section 3 as an illustration of some of the issues to be addressed
for coastal ocean forecasting in this situation. The data synthesis aspects of
the modeling, not yet realized, are quite different from CBLAST. Firstly,
variational data assimilation (Moore et al. 2004, Weaver et al. 2003) applied
to the measured dye distribution will be used to provide high a resolution
hindcast of the evolving plume trajectory and physical structure. Secondly, a
coupled physical/biological model that explicitly computes the bio-optical
features of suspended and dissolved matter (Bissett et al. 1999a, 1999b) will
be used to hindcast the depth-dependent distribution of phytoplankton and
DOM and predict the inherent optical properties of the water column. This
JOHN L. WILKIN AND LYON LANEROLLE