Several platforms were prepared for surbsurface sampling during the 1998 field season' Rutgers physi

1
Adaptive Sampling of Coastal Upwelling at the
Long-term Ecosystem Observatory -- LEO-15 Scott
Glenn, Michael Crowley, Frederick Grassle, Dale
Haidvogel, Oscar Schofield, Institute of Marine
and Coastal Sciences, Rutgers University, New
Brunswick, New Jersey Christopher von Alt, Ocean
Systems Laboratory, Woods Hole Oceanographic
Institution, Woods Hole, Massachusetts
Real Time Monitoring
Coastal Upwelling Historical Perspective
New Subsurface Observations
Adaptive Sampling Strategy
The strong southerly winds
responsible for the surfacing of the thermocline
on July 17 continued through July 20,
causing the newly formed upwelling front
to move further offshore. After the winds
relaxed on July 20 to light and variable for
several days, explosive growth of the
upwelling center was observed. By the morning
of July 23, a fully developed upwelling
center with an embedded cyclonic eddy was
observed in the real-time AVHRR-derived sea
surface temperatures and CODAR-derived
surface currents. The physical survey
vessel equipped with the towed ADCP and
CTD/ Fluorometer systems was dispatched
to sample the northern cross-shelf
transect (black line). The Survey REMUS AUV was
sent to sample along a southern
cross-shelf transect between Node B and
Mooring C (red line). By mid-day, Turbulence
REMUS was sent along an intermediate
cross-shelf line extending from the
center of the cyclonic eddy across the
upwelling front (white line).
Velocity 20 cm/sec
LEO-15 is the site of numerous
multi-institutional inter-disciplinary research
efforts that have involved over 50 researchers
from over 20 institutions. The
extensive observational infrastructure
has provided researchers interpreting
their own observations a greater physical
context than is typically available.
Real-time access to the remote sensing,
meteorological and LEO-15 monitoring data
has proven itself to be a critical
mission planning device for scientists
directing shipboard and AUV sampling patterns.
NOPP support in 1998 resulted in the
addition of several new real-time products to
our Internet Web site
(http//marine.rutgers.edu/mrs) to support the
summer 1998 field program. Web access
doubled during 1998, averaging over 20,000
hits per day during July, with 70 coming
from commercial Internet service providers.
Web-based products developed and used in a
research mode for NOPP are then transitioned to
operational status and maintained by the
Mid-Atlantic Bight National Undersea Research
Center.
Typical summer Bermuda Highs generate
large- scale southerly winds in the Middle
Atlantic Bight, resulting in the upwelling of
cold, nutrient rich water along the New
Jersey coast. Rather than forming a
continuous band of cold water at the surface as
expected in 2-D situations, the underlying
topography causes the formation of three
recurrent upwelling centers along southern
New Jersey (left). The coldest water within
each of the upwelling centers is co-located with
the previously observed regions of recurrent
hypoxia . OSCR-derived surface currents (below)
at LEO-15 reveal the formation of a strong
alongshore jet on the warm side of
the upwelling front, and a cyclonic
eddy within the upwelling center.
in
Observations of the sub-inertial
sub-surface flow field associated with an
upwelling center requires removal of the tidal
and near-intertial oscillations in the
ship/AUV ADCP data. To resolve these
spatially-varying oscillations, survey patterns
were repeated along fixed transects for
two-day periods. A typical sampling pattern
required the Rutgers survey vessel to conduct
repeat transects along one 20 kilometer
cross-shelf line while the WHOI Survey REMUS
conducted repeat transects along an adjacent
20 kilometer cross-shelf line. The NUWC
Turbulence REMUS was then free to
roam on a dedicated mission between these
two transects.
21 22 23 24 25 26 27
The upwelling centers also are observed
to contain enhanced concentrations of
phytoplankton, which can deplete the
dissolved oxygen as the dying plankton
settle and decay. But where is the source
of the cold water that contains the
phytoplankton? Is the cold water coming in
directly from offshore as expected in the
2-D case, or is it coming in from the south
similar to the observed surface jet? Can the
phytoplankton be trapped in the cyclonic eddy
long enough to affect oxygen levels?
Observations of the subsurface flow field
within an upwelling center are required to
begin to address these questions.
Mon Tues Wed Thur
Fri Sat Sun Mon
Tue Wed Thur
An ocean forecasting system was
assembled to guide the design of the
survey pattern for each two- day period.
Because the two-day ADCP surveys were
conducted twice per week on Monday/Tuesday
and Thursday/Friday, forecast model runs were
targeted for submission on Sundays and
Wednesdays. The model runs would assimilate
remote sensing surface and shipboard
subsurface data up to this cutoff. The
4-day forecasts would run overnight and be
available the next morning for mission planning.
Marine Remote Sensing Lab homepage
displaying selected icons for accessing
datasets.
LEO-15 Subsea Nodes
The LEO-15 system consists of two
instrument platforms (or nodes) housed within
trawl-resistant steel shells that are anchored
to the sea bottom. The nodes are equipped
with video cameras, acoustic hydro- phones,
stationary tide, wave and current sensors,
instrumented vertical profilers, and 8 guest
ports to plug in additional sensors.
Marine Remote Sensing Lab homepage average daily
hits for 1995-1998
Adaptive Sampling Platforms
Two concurrent parallel velocity sections
obtained from the SWATH towbody and the REMUS AUV
are displayed above, with the alongshore
velocity displayed as color-coded contours and
the cross-shore velocity indicated with
arrows. On the offshore side of both sections,
the previously observed northward flowing surface
jet is located on the warm side of the upwelling
front. On the inshore side, a new subsurface
feature was discovered. In the final 4
kilometer range cells for both sections, a
strong southward flowing subsurface jet is
observed in the alongshore flow within the eddy.
Several platforms were prepared for
surbsurface sampling during the 1998 field
season. Rutgers physical survey vessel (R/V
Caleta) was equipped with two new towed
instrument systems. An ADCP mounted on an
aluminum Small Water Area Twin Hull (SWATH)
vehicle was towed alongside the Caleta, and
a CTD/Fluorometer mounted on an undulating
vehicle was simultaneously towed behind Caleta.
The WHOI docking port for
the REMUS AUVs was connected to LEO Node A for
docking tests, and the WHOI long-range (up to 60
kilometers) Survey REMUS AUV, equipped with
upward and downward looking ADCPs and
a CTD/Fluorometer, was ship deployed along an
acoustic navigation line starting at Node
B. NUWC provided the
Turbulence REMUS AUV equipped with several
microstructure sensors mounted on
stings extending beyond the vehicle nose. WRC
deployed a Lagrangian Multi-Trip CTD in
preparation for the summer 1999 deployment of
their SLOCUM Glider AUVs equipped with
CTD/Fluorometers and designed for month-long
patrolling missions offshore.
New York Bight real-time SST satellite
imagery (one of fourteen areas on the web site)
The numerical model used to generate
the ocean forecasts was the Rutgers Regional
Ocean Modeling System (ROMS v1.0). Data
assimilation was via a multidata-variate
Optimal Interpolation (OI) scheme. Surface
forcing was provided by operational Navy
atmospheric forecasts (NORAPS and NOGAPS). The
most successful of the ocean forecasts was the
prediction of the initial surfacing of the cold
upwelled water at LEO-15 on July 17 during
persistent large-scale southerly winds accurately
portrayed in the Navy atmospheric forecasts.
The vertical profiler can be programmed to
cycle autonomously or can be remotely
operated from the shore station or over the
Internet. The instrument package as well as the
vertical profiler are retrievable for routine
servicing and upgrades.
Observational Network Surrounding LEO-15
Real-time detided and filtered CODAR surface
velocity fields during July/August
Temperature/Conductivity Profiles from Node B
Surface Towed SWATH ADCP
Undulating CTD REMUS Docking
Port
Survey REMUS AUV
Turbulence REMUS AUV
Webb Glider CTD
Concurrent CTD/Fluorometer sections also
were acquired with the undulating towbody along
the northern transect. The upwelling front is
observed on the offshore side of the temperature
section at a range of 8 kilometers. On the
inshore side between 24 and 28 kilometers, the
isotherms spread vertically possibly due to
mixing by the strong subsurface jet. The
corresponding fluorometer section along this
transect reveals extremely high phytoplankton
concentrations nearshore below the thermocline,
and virtually no signal offshore above the
thermocline. The subsurface data acquired from
the adaptive sampling surveys gave us our
first indication that the source of the
phytoplankton within this upwelling center is
this previously unknown subsurface jet flowing
southward nearshore below the thermocline.
The coastal adaptive sampling
network centered on the LEO-15 site was
developed through several partnerships
funded mainly by ONR and NOPP. The
observation network includes three satellites
(AVHRR for sea surface temperature, SeaWiFS for
ocean color, RADARSAT for surface roughness), a
shore-based CODAR for surface currents, a
meteorological tower for local forcing and
offshore RF communications, numerous
moored sensors for time series, and several
Research Vessels and AUVs for subsurface
spatial sampling. Recent additions
include the hyperspectral NEMO satellite
and AVIRIS aircraft, meteorological
profiles (SODAR), and
several meteorological and
radiometric buoys.
Real-time wave heights updated every 15 minutes
Acknowledgements
Time Series of Filtered Cross Shore Velocity
(cm/sec)
Sponsors of this project include The
National Science Foundation (NSF), the Office of
Naval Research (ONR), NOAAs Mid-Atlantic Bight
(MAB) National Undersea Research Center (NURC),
and the National Oceanographic Partnership
Program (NOPP), a collaboration of twelve
federal agencies to promote cooperative
activities among government, academia, and
industry, for the advancement of ocean science,
technology and education. NOPP Partners
include Ed Levine (NUWC), Don Barrick and
Belinda Lipa (CODAR), Joel Young (RDI), Doug
Webb (Webb Research Corporation), Roni
Avissar (Rutgers), and Rich Signel
(USGS). Additional partners include Stan
Schneider and John Duchesney (Falmouth
Scientific), John Joynt (Guildline) and
David Walworth (Gold Coast Yachts).
Rutgers researchers include Hernan Arango, Robert
Chant, Liz Creed, John Fracassi, Rose Petrecca
and John Wiggins. Woods Hole researchers include
Tom Austin, Roger Stokey, Ben Allen, Ned
Forrester, Rob Goldsborough and Mike Purcell. The
1996 data was collected through an NSF-sponsored
project with Rich Garvine (U. Delaware) and
Andreas Munchow (Rutgers).
Weekly wind data updated every 15 minutes
Time Series of Filtered Along Shore Velocity
(cm/sec)
05 06 07 08 09 10 11 12 13 14 15 16
17 18 19 20 21 22 23 24 25 26 27 28
29 30 31
The adaptive sampling experiment ran for
4 weeks starting July 6, 1998. Available remote
sensing data includes 18 days of clear AVHRR
coverage, 10 RADARSAT overpasses, and a pair of
AVIRIS aircraft overflights on July 12. Shipboard
cruises making use of the real-time datasets and
model forecasts for adaptive sampling includes
the physical, bio-optical, chemical and
zooplankton surveys. The Survey REMUS AUV logged
261 km on 10 deployments, and the Turbulence
REMUS was successfully deployed 4 times.
Weekly detided and filtered Node A ADCP
Velocities
Real-time video from Node B