American Meteorological Society 81st Annual Meeting
Fifth Symposium on Integrated Observing Systems
Rutgers University Institute of Marine and Coastal Sciences Presentations

1.2
DEVELOPMENT OF A REGIONAL PHYSICAL/BIO-OPTICAL COASTAL OBSERVING SYSTEM FOR RAPID ENVIRONMENTAL ASSESSMENT IN THE NEW YORK BIGHT.
Scott M. Glenn Rutgers University, New Brunswick, NJ 08901., Oscar M.E. Schofield, Dale B. Haidvogel, and J. Frederick Grassle.

The Rutgers University Long-term Ecosystem Observatory (LEO-15) located offshore Tuckerton, New Jersey will host the third in an on going series of Coastal Predictive Skill Experiments in July 2000. The annual NOPP/ONR supported CPSEs focused on improving nowcast skill via data assimilation in July 1998, and on improving forecast skill through improved boundary conditions and turbulent closure in July 1999. The purpose was to provide guidance for the adaptive sampling with ships and Autonomous Underwater Vehicles (AUVs) of the recurrent upwelling centers that occur each summer along the southern New Jersey coast. The July 2000 CPSE will further test Rapid Environmental Assessment techniques to guide the physical and bio-optical sampling of these coastal upwelling centers, the adjacent Mullica River estuary, and the freshwater plume from the Hudson River. Over 170 scientists, students and researchers from over 20 institutions will participate.

Rapid Environmental Assessment methodologies use a combination of model forecasts and real-time datasets to optimize sampling scenarios. The forecasting system for July 2000 includes both operational Navy and locally-run high-resolution atmospheric and oceanic models for the atmospheric forcing and the coastal ocean response, bottom boundary layer models for sediment transport, and radiative transfer models for remote sensing reflectance. Real-time datasets include an expanded constellation of satellite sensors (AVHRR, SeaWiFS and FY-1C), surface currents from HF-Radar, subsurface data from ships and a cross-shelf array of moorings transmitted to shore via Freewave modem, and meteorological forcing data from shore-based towers/SODARs and offshore buoys. The forecasts and real-time data will guide an expanded physical/bio-optical adaptive sampling network including ships, AUVs, aircraft and divers. The new aircraft-based sensors include the hyperspectral PHYLLS (NRL) and AVIRIS (NASA) sensors for high resolution ocean color, the Microwave Salinity Mapper (NRL) to map the Hudson river plume, two aircraft altimeters (JHU/APL) to map the sea surface height, and an airborne LIDAR (NAVAIR) to map the subsurface optical structure. Several of the observation systems to be tested during the July 2000 CPSE will be used to construct a regional observatory for the New York Bight (LEO-NYB). An overview of CPSE highlights will be presented.

1.3
MULTIPLE HF-RADAR SYSTEM DEVELOPMENT FOR A REGIONAL LONGTERM ECOSYSTEM OBSERVATORY IN THE NEW YORK BIGHT
Josh T. Kohut Rutgers University, New Brunswick, NJ 08901., Scott M. Glenn, and Donald E. Barrick.

A Standard SeaSonde HF-Radar system has been operational at the Longterm Ecosystem Observatory (LEO-15) along the New Jersey coast since 1998. The system provides real-time maps of ocean surface currents extending 50 km alongshore and 40 km offshore. The various in situ measurements, including moored, ship-towed and AUV mounted ADCP’s, collected at LEO-15 provide an excellent testbed for HF-Radar validation studies. An important aspect of this validation has been to test the role of antenna pattern distortions in both the accuracy and coverage of the measurements. Since the antenna patterns are used by the system to determine the direction of the backscattered signal, any distortions in these patterns will affect the surface current data. An experiment has been set up to test possible causes of these distortions including the surrounding environment and system hardware. By optimizing the system based on the antenna patterns, the radial data collected has been shown to be more accurate and cover a larger area of the ocean surface. In addition to the standard SeaSonde system, a long-range system was deployed in June 2000. This new site transmits at a lower frequency allowing measurements to be made further offshore. Radial data collected during the summer of 2000 will be validated against multiple ADCP platforms at the LEO-15 site. One of the restrictions of HF-Radar systems is the inability of the systems to resolve total current vectors near the coast. The third system to be tested at LEO-15 will eliminate this near coastal gap by incorporating data collected by two bistatic transmitters mounted on buoys deployed 20 km offshore. The standard, long-range and bistatic CODAR systems form the backbone of a nested regional observatory for the New York Bight (LEO-NYB). LEO-NYB is one prototype for a series of linked regional observatories envisioned to form the NorthEast Ocean Observing System (NEOOS).

1.5
FLOW STRUCTURE AND MULTIPLE-SCALE TOPOGRAPHY IN THE COASTAL OCEAN.
Robert J. Chant Rutgers University, New Brunswick, NJ 08901., Scott Glenn, and Philip Bogden.

Coastal upwelling has been the focus of both modeling and observational efforts on New Jersey's inner shelf as part of the NOPP/ONR funded efforts in the vicinity of Rutgers University's LEO-15. Particular effort has focused on an upwelling center which develops as winds relax immediately to the north of a topographic high. The spin-up of the upwelling center appears to be related to a strong near-shore upwind jet which veers offshore as in impinges on the topographic high and feeds the upwelling center with cool water. While the both the topography and the upwelling center have a common 25-50 km length scale, superimposed on this larger topographic scale is smaller scale associated with oblique sand ridges. These sand ridges modulate water column depths by up to 25 percent and have a horizontal length scale 5-10 km. The ridge/swale topography is significantly more pronounced over the topographic high. Thus the multiple-scale topography along the New Jersey coast is characterized by relatively smooth topographic lows and significantly rougher topographic highs. Over the past three years shipboard and autonomous ADCP surveys have completed numerous repeated transects over both topographic features. The influence of this multiple-scale topography on detided depth averaged and depth dependent flows are discussed based on a statistical description of the entire data set and on event scale flow features.

1.9
INTEGRATION OF THE INTERNATIONAL REAL-TIME OCEAN COLOR DATA TO THE NEW JERSEY LONG TERM ECOSYSTEM (LEO-15)
Oscar Schofield Rutgers University, New Brunswick, NJ 08901., Trisha Bergmann, Michael Crowley, and Scott Glenn.

Our efforts have focused on the development, demonstration and evaluation of an integrated adaptive sampling and modeling system for nowcasting and forecasting the 3-dimensional evolution of the physical and optical properties in the nearshore coastal ocean. Optical data is used to provide inputs to a suite of newly developed ocean color algorithms providing estimates of both biological (phytoplankton, particulate organic carbon) and chemical (colored dissolved organic matter) material. The optical data at LEO-15 is collected by a flexible array of in situ assets (including ships, robotic moorings and autonomous AUVs). During the 1999 summer experiments the impact of coastal upwelling on bulk apparent and inherent optical properties was quantified and related to phytoplankton biomass and composition. The measured optical properties were used as inputs to the Hydrolight radiative transfer model (RTE) and provided estimates of remote sensing reflectance. These model estimates were compared to satellite-derived estimates. Quantitative agreement between the SeaWiFs-measured and in-water modeled remote sensing reflectance was good, but results were variable depending on the specific reflectance model being used. The ocean color satellite data however, was limited by the revisit schedule of the SeaWiFs ocean color satellite. By accessing data from the growing constellation of ocean color satellites there is the potential to increase the observations of biological and chemical constituents on a time scale more commensurate with current hydrographic/atmospheric observations. To this end during the summer 2000 experiments, we began collecting data from both SeaWiFs and the Chinese FY1-C in order to minimize site revisit intervals and provide coverage in both the morning and afternoon. The real-time data allowed for optimization of field sampling efforts. In turn, the collected optical data was used to validate and cross-calibrate the products from both satellites.

1.11
RODAN: RUTGERS OCEAN DATA ACCESS NETWORK POWERED BY JAVA TECHNOLOGIES
Yunqing P. Zhang Rutgers Univ., New Brunswick, NJ 08901., John F. Fraccassi, John E. Wiggins, Scott M. Glenn, and Frederick Grassle.

With the rapid development of remote sensing and observation technologies, scientists have been able to collect an unprecedented amount of data on our ocean environment in the past decade. Strategic and structural data management is essential in successfully transforming this vast reservoir of information into a useful scientific data product. The Long-term Ecosystem Observatory (LEO), located offshore of Tuckerton, New Jersey, is composed of two instrument platforms (nodes) at 15 meter depth and a surrounding network of satellite, aircraft and shore-based remote sensing systems (Grassle et al., 1998; Glenn et al., 1999). The major purpose of the Observatory is to obtain oceanic and meteorological observations over broad temporal and spatial scales. To facilitate open access and fast distribution of archived LEO data, we have established the Rutgers Ocean Data Access Network (RODAN), an integrated data management and analysis system. RODAN supports customized search, retrieval, analysis and visualization of data through an intuitive web interface. Using the widely-hailed three-tier client-server technology, RODAN integrates and streamlines the previously separated tasks of storing, retrieving, searching, analyzing and plotting data. In addition, the new multi-tiered and multi-threaded servlet framework of RODAN decreases database connection time and ensures scalability for future growth. Moreover, multithreaded server applications allow RODAN to perform multiple tasks simultaneously, maximizing efficiency. As a result, the adoption of state-of-the-art Java 2, especially servlet and 2-D technology-based Application Programming Interfaces (APIs), makes RODAN more robust and universally transferable. Since its inception, RODAN has been providing data services to numerous users from universities and research laboratories and has become a well-regarded venue for scientific exchange and collaboration.

2.7
VALIDATION DATASETS FOR COUPLED REGIONAL ATMOSPHERE/OCEAN MODELS.
Katherine S. Hedstrom, Rutgers University, New Brunswick, NJ; and D.B. Haidvogel

There are at present within the field of ocean general circulation modeling four classes of numerical models which have achieved a significant level of community management and involvement, including shared community development, regular user interaction, and ready availability of software and documentation via the World Wide Web. These four classes are loosely characterized by their respective approaches to spatial discretization (finite difference, finite element, finite volume) and vertical coordinate treatment (geopotential, isopycnic, sigma, hybrid). Given the rapidly growing number of models, and their algorithmic options, it is necessary that we understand the relative behavior of these models and their component methods. Several alternative approaches to model characterization are possible. One of these, based upon an inexpensive set of process-oriented test problems, is the basis for an ocean model test problem web site recently created by the authors (http://marine.rutgers.edu/po/tests).

Another possibility is the assembly of a parallel suite of model test problems based on intensive geophysical datasets -- e.g., as collected in the New York Bight at the LEO-15 National Littoral Laboratory. As of summer 1999, the modeling infrastructure at LEO-15 utilized the latest version of our coupled regional ocean/atmosphere model (ROAMS), whose combined components include a high-resolution regional ocean circulation model and a nested mesoscale meteorological model run at 4 km local resolution. Physical processes addressed at LEO-15 include frontal development, air-sea coupling, the effects of stratification and topography, and vertical mixing in the coastal ocean.

We describe progress in creating a high-resolution coastal prediction skill assessment database derived from the real-time activities at LEO-15 during the summer of 2000. The goal is to provide a uniformly gridded, high-quality suite of initialization, surface forcing, and verification data with which to assess and to compare alternative coupled atmosphere/ocean circulation models, including approaches to data assimilation. Examples of forecasts from the ROAMS model are provided as an example of current forecasting capabilities.

2.11
ADVANCED ASSIMILATION STRATEGIES IN MODERN OBSERVATIONAL NETWORKS FOR REAL-TIME, HIGH RESOLUTION APPLICATIONS
Hernan G. Arango Rutgers University, New Brunswick, NJ 08901-8521., Pierre J. F. Lermusiaux, and Scott M. Glenn.

Rapid advances in computer technology and estimation/simulation algorithms over the last few years have made possible the melding of multiplatform observations with models in real time using advanced and versatile assimilation methodologies. The error subspace statistical estimation (ESSE) scheme is used to assimilate the various data gathered by the observational network at the Long-Term Ecosystem Observatory (LEO-15) into a coupled atmosphere-ocean (RAMS/ROMS) forecast system along the Southern New Jersey coast. The ESSE is a nonlinear, multivariate minimum variance approach based on the dominant optimal reduction of the dimension of the time-variant, state error covariance matrices. A Monte Carlo perturbation ensemble method is used to determine the evolution and adaptive learning of the dominant variability and error covariances. Up to 300 perturbations, per assimilation cycle, are needed in this application (with a state vector of about 2.75 million points) to achieve convergence of the error subspace forecast. Massive parallel computers are used to assimilate new data, predominantly surface velocity data from CODAR, every six hours. The error subspace approach is used to determine the adaptive sampling and better use of the observational assets.

Version 1.3 Released February 14, 2001 This CD was produced by Rutgers Coastal Ocean Observation Lab Please e-mail all inquiries to: flounder@arctic.rutgers.edu. © 2001 RU COOL