The New Jersey Shelf Observing System

Transcription

1 The New Jersey Shelf Observing System Scott M. Glenn and Oscar M.E. Schofield Institute of Marine and Coastal Sciences Rutgers University 71 Dudley Road New Brunswick, NJ ABSTRACT-The New Jersey Shelf Observing System is a coastal ocean observatory whose primary goal is supporting collaborative interdisciplinary oceanographic research. The observatory has both a sustained component designed to provide spatial datasets yearround, and a process study component for more intensive measurements during shortterm scientific experiments. The sustained component consists of tracking stations for the international constellation of ocean color and IR satellites, multi-frequency multistatic CODAR HF Radars, and long-duration subsurface Glider AUVs. The processes study component uses numerous platforms that include aircraft, ships, propeller-driven AUVs and relocatable mooring arrays. Process studies focused on recurrent coastal upwelling centers and their biological impacts from , and are planned to focus on the Hudson River plume, chemical contaminants, and their biological impacts from Despite being a research-oriented observatory run by the scientists for the scientists, it maintains a significant societal impact through its website (marine.rutgers.edu/cool), receiving an average of over 60,000 hits/day during the busy summer months. I. INTRODUCTION The New Jersey Shelf Observing System (NJSOS) is an evolving research-oriented coastal ocean observatory [1,2,3]. As a research observatory, a primary goal of NJSOS is the development of the systems and capabilities required to provide scientists a real-time data-rich environment in which to conduct more detailed process studies. This concept was developed during a series of interdisciplinary Coastal Predictive Skill Experiments (CPSE) conducted offshore Tuckerton, New Jersey each summer from [4,5,6]. It has proven its ability to attract numerous researchers to field programs well beyond those funded in the original grants [7]. Still Time-series Observation Sites PA 75W RU Field Station LEO Node 74W Long Island Ambrose Light 40N 39N Weather Station NOAA buoy Fig. 1: Existing time-series observation sites in the New York Bight. the research observatory maintains a significant societal impact. Documented users include the Coast Guard for search and rescue, NOAA for oil spill response and weather forecasting, NAVO for homeland defense, NJDEP for harmful algal blooms, and a vast number of fisherman and recreational interests. Adopting the language of the Ocean.US report on a Sustained and Integrated National Ocean Observatory, NJSOS is a regional enhancement of an existing national backbone of long-term observation sites. Several NOAA weather buoys and tide stations are distributed throughout the New York Bight (NYB) region (Fig. 1). In addition, the NOAA Mid-Atlantic Bight (MAB) National Undersea Research Center (NURC) maintains Rutgers Long-term Ecosystem Observatory (LEO-15), a seafloor cabled observatory located offshore Tuckerton. These systems are designed to provide long-term time series at points on the New Jersey continental shelf. The Ocean Sciences Decadal Committee applauds their progress, but urges us to do more over the coming decade. Their report states: The very few existing time-series stations paint a

2 Fig. 2. New Jersey Shelf Observing System compelling picture of important oceanic changes in physics, chemistry and biology. Yet these stations capture the time domain at only a single point. New strategies for observing the appropriate spatial correlation are required (Ocean Sciences at the New Millennium, 2001). NJSOS (Fig. 2) fills in the spatial gaps between the existing point measurements. It has a sustained component designed to provide year-round continuous synoptic coverage of the New Jersey shelf. It also has a process study component designed to provide intensive coverage in smaller subdomains during short-term scientific experiments. II. SUSTAINED COMPONENTS OF NJSOS The three sustained components of NJSOS are: (1) data acquisition systems tracking an international constellation of IR and ocean color satellites [3], (2) a nested grid of long-range (5 MHz) and high-resolution (25 MHz) multi-static CODAR HF Radars [8,9], and (3) a growing fleet of long-duration autonomous underwater Gliders [10,11]. Satellites provide real-time maps of sea surface temperature, phytoplankton and suspended sediment distributions. The CODAR Radar array provides real-time maps of the surface current distributions every 1-3 hours over large portions of the continental shelf. The Glider fleet patrols beneath the surface maps to provide vertical sections of temperature, salinity, phytoplankton and sediment distributions and depth-averaged currents. The datasets are routinely assembled into data-based nowcasts of present conditions, and have been assimilated into models to generate forecasts of future conditions. Results are distributed via the World Wide Web to numerous users. For scientists, the real-time nowcasts and forecasts provide the information necessary for adaptive sampling [12] with the Gliders and other mobile platforms. International Constellation of Satellites. Earth observing satellites often use direct broadcast capabilities to transmit full resolution data to anyone operating a line-of-site ground receiving station, either in the traditional L-Band or in the new higher baud rate X-Band. Rutgers has operated an L-Band satellite data acquisition system continuously since The system currently provides real-time access to the full-resolution direct-broadcast raw data streams from the AVHRR sensors on four NOAA satellites, the SeaWiFS ocean color sensor, and ocean color sensors on China s FY1-C and FY1-D. Tracking the full constellation of satellites with different overpass times reduces revisit intervals. Multiple satellite passes in a day provide researchers with the ability to track rapidly changing features on daily scales and to work around diel weather patterns (such as June 4, 2000 Chlor-a mg m Wind June 11, 2000 Chlor-a mg m m Wind 40N m 40m 60m 40N July 6, 2000 Chlor-a mg m Wind m 60m 40N 100m 100m 100m 39N 39N 39N 74W 73W 74W 73W 74W 73W Fig. 3. The evolution of a surface plume from the Hudson River in the summer of 2000 mapped using SeaWiFs.

3 TABLE 1 Current and anticipated earth observing systems with X-Band direct broadcast capabilities. Satellite/ Sensor RADARSAT SAR IRS-P3 MOS IRS-P4 OCM EOS Terra MODIS EOS Terra MODIS ENVISAT MERIS HY-1 COCTS/CCD Orbview-3 Pan & MSI ADEOS-2 GLI RADARSAT2 SAR NEMO COIS Nation/ Bands Canada 7 India 17 India 8 USA 36 USA 36 Europe 17 China 14 USA 5 Japan 36 Canada 2 USA 211 Resolution Launch (km) Date 27 In orbit In orbit 0.35 In orbit In orbit In orbit.3 In orbit.25-1 In orbit Fall Nov Spring Spring 2004 afternoon cumulus) that can compromise untimely satellite passes. Real-time local access enables adaptive sampling strategies to be implemented by allowing researchers to choose which products are produced first from the late morning passes. These priority products can then be used to position ships and AUVs for early afternoon passes. New satellite algorithms being developed by ONR s HyCODE program have focused on deriving phytoplankton, suspended sediments, and CDOM in turbid coastal waters [13,14]. For example, Fig. 3 illustrates a series of SeaWIFS-derived Cl-a estimates showing a plankton plume whose surface signature evolution is consistent with an upwellingfavorable wind-forced buoyant plume. X-Band satellite data acquisition systems are capable of acquiring data from the new generation of higher resolution (both spectral and spatial) ocean color satellites (Table 1). New satellites with X-Band direct-broadcasting capabilities most easily accessed include the U.S. MODIS (Terra and Aqua), India s Oceansat, China s HY1, and Japan s planned ADEOS-II. Higher spectral resolution enables users to better identify phytoplankton, CDOM and suspended sediment in complex coastal waters, and higher spatial resolution allows one to better resolve sharp nearshore fronts [15]. Real-time local access again enables researchers to more efficiently use the data for adaptive sampling by providing high quality maps of what is in the water. The next need is to define where the material is being transported. CODAR HF Radar Network Rutgers operates an array of ten CODAR HF Radar systems that include six long-range (5 MHz) backscatter systems, two high-resolution (25 MHz) backscatter systems, and the first two buoybased bistatic transmitters (5 MHz and 25 MHz). Four of the long-range CODARs have been deployed along the New Jersey coast since spring of 2001, providing coverage of the entire shelf at 6 km resolution every 3 hours (Fig. 4). The two 25 Mhz CODARs have been deployed since 1998, providing enhanced resolution (1.5 km every hour) in the vicinity of Tuckerton. The full array is currently being operated in monostatic mode, where each site only listens for the backscatter from its own transmitted signal. All systems are being upgraded with GPS timing to run Fig. 4. Real-time Long-range CODAR Coverage. Fig. 5: 25 MHz Bistatic Buoy Transmitter.

4 in multistatic mode, enabling each site to listen to scattering from all sites within its operating range [9,3]. By adding this capability, N monostatic radars become N 2 multistatic radars, virtually eliminating traditional Geometric Dilution of Precision (GDOP) concerns offshore. However, even a multistatic shore-based HF Radar system will have difficulty measuring flows leaving and entering an estuary since the current components measured nearshore are approximately aligned with the coast, resulting in large GDOP nearshore. Bistatic transmitters deployed offshore on buoys (Fig. 5) are specifically designed to provide current component estimates perpendicular to the coast to significantly reduce GDOP, thereby increasing the effectiveness of HF Radars for measuring near-shore flows, particularly in and out of estuaries. Combining the CODAR surface current maps with the satellite imagery then provides researchers with synoptic spatial estimates of what Fig. 6: Satellite SST image of a recurrent upwelling center with overlaid CODAR surface currents showing the upwelling eddy. is in the water and where it is going. Fig. 6 is one example of this. The map is of a recurrent upwelling center often observed offshore Tuckerton, NJ. The satellite image displays the strong surface front that develops, and the overlaid CODAR vectors illustrate the associated cyclonic flow field. Real-time access to these maps have proven their value for coupled physical/biological adaptive sampling in previous Coastal Predictive Skill Experiments [16,17]. But these surface maps also need to be augmented with subsurface data. Autonomous Underwater Glider Fleet. Rutgers currently operates a fleet of 4 Slocum Electric Glider AUVs [11]. The Gliders (Fig. 7) are equipped with CTDs, and are being upgraded with different optical sensors (backscatter/fluorometer systems or hyperspectral fluorometers). The Gliders are designed for flights along sections through regions of enhanced phytoplankton growth often observed in satellite imagery. The optical sensors will help validate and calibrate satellite imagery. Gliders with these instrument payloads are capable of month-long deployments with standard alkaline battery packs. They undulate through the water column and surface every hour to transmit data and receive new navigation and sampling commands. Depth averaged currents are calculated each time the Glider surfaces by differencing its new GPS location from its dead-reckoning position. Typical forward speeds of the Gliders are 35 cm/sec, or 30 km/day relative to the water. Typical vertical speeds are 12 cm/sec. Freewave radio modems are used for high baud rate line of sight communications through repeaters located on tall nearshore structures and on aircraft. Iridium is used for low baud rate global communications. A Glider mission control center is currently Fig. 7: Slocum Electric Glider AUV. Fig. 8: Glider Mission Control Center demonstration of the World Wide Web display for Glider locations.

5 being constructed to coordinate operation of a smart Glider fleet using an agent oriented programming approach developed for selfaware/self-controlled robots. This is similar to software used by NASA's Deep Explorer spacecraft to navigate itself through the tail of Comet Borrelly. Software agents that enable the Gliders to assume their own vertical control have already been written and will be tested in the New York Bight and on the West Florida Shelf. The software agents interpret data obtained from the Glider's sensors, and using its growing knowledge base, decide what region of the water column to sample. Examples include undulations above or below a thermocline, or following a subsurface phytoplankton peak. Software agents for horizontal control will be constructed based on West Florida Shelf s red tide data (track-a-bloom agents) and the Hudson River Plume data (track-a-plume agents). Development of a society of software agents for control of the Glider fleet is also underway in collaboration with Dr. Naomi Leonard (Princeton). This includes transitions between standard cross-shelf transects and Dr. Leonard's different swarming behaviors when specific features are detected. Fig. 9: High-resolution hyperspectral image of nearshore sediment plumes observed with the NRL PHILLS aircraft. III. PROCESS STUDY COMPONENTS OF NJSOS The process study component of NJSOS consists of relocatable systems on aircraft, ships, autonomous underwater vehicles (AUVs) and moorings that can be assembled to study specific scientific questions within the well-sampled domain of NJSOS. These platforms are usually operated for time periods measured in hours to weeks rather than year-round, and often have limited sampling range. It therefore is critical to have real-time data available to coordinate where and when the different systems will be deployed. Examples of this approach include the summer physical/biological studies of a recurrent coastal upwelling center offshore Tuckerton (Fig. 6), or planned spring studies of the Hudson river plume as it flows onto a stratified, wind-driven shallow shelf (Fig. 3). Aircraft provide access to hyperspectral ocean color imagery at much finer spatial scales than the satellites. Increased spatial resolution is important nearshore, since many processes occur on spatial scales much less than the usual 1 km pixel resolution available from satellites with global missions. Fig. 9 is but one example of nearshore sediment plumes observed at the 2-meter resolution available from an aircraft sensor. In addition, aircraft provide an excellent platform for radio modem repeaters used to communicate with the Glider AUVs. Since local radio modem Fig. 10. Towed CTD/Fluorometer systems mounted on an undulating Mini-bat platform. communications are orders of magnitude faster than global Iridium satellite communications, even short aircraft flights enable downloading of complete mission data files from the Gliders. Iridium communications, because of their expense, are typically limited to transmission of a small subset of critical science parameters. Typically three research vessels form the core fleet for dedicated process studies. Shipboard access to the NJSOS datasets is currently provided via Freewave radio modems using repeaters placed on tall coastal structures [18]. Since radio modem range is highly dependent on the height of the lowest antenna, tall ship superstructures increase the range well beyond that achieved by the sea level antennas on board the Glider AUVs. Usually a traditional small UNOLS vessel suitable for coastal work is used for continuous 24-hour operations for durations on the order of a week to 10 days. Vessels used in the past include the R/Vs Endeavor and Cape Henlopen. This provides a large working platform for mooring and tripod deployments, large capacity profiling systems, and offshore biological labs for fast turn around of water samples. Two smaller day-boats capable of very shallow water operations are used for faster response to changing oceanographic conditions.

6 One with small towed instrumentation and rugged sensors (Fig. 10) is used for fast fine-resolution coverage of larger areas. Real-time transmission of data collected by this boat enables it to also be used as a scout vessel, leaving the dock early in the morning well before the sun is high in the sky for optimal bio-optical measurements. The second coastal vessel is equipped with more extensive physical/bio-optical profiling systems with more sensitive sensors and water samplers, enabling detailed studies in specific areas not easily accessible by a large research ship. AUVs used within NJSOS include the propeller-driven REMUS vehicles designed for flying short duration (measured in hours) high power consumption missions. REMUS vehicles are small and lightweight, making them ideal for hand deployment by a single person on a small coastal research vessel. They are navigated with an acoustic network that also is easily deployed and recovered from small vessels for durations of days to a month. REMUS AUVs provide extremely stable subsurface platforms with the flexibility to attach different payloads, including sensors for bathymetry, currents, temperature, salinity, bioluminescence and turbulence. With a maximum speed of 4 knots and durations measured in hours, it is important to deploy the REMUS vehicles in the right place at the right time to acquire the desired measurements. In 2000 and 2001, a REMUS equipped with a bioluminescence sensor (Fig. 11) was used for nighttime sampling of convergence zones observed in the real-time CODAR data. The convergence zones were often only a few kilometers wide, and would be difficult to locate without real-time guidance. Lastly, a relocatable mooring array is available for typically 6-week deployments in regions of scientific interest. The array is specifically designed with inexpensive sensors so that several moorings can be constructed. Typically each mooring consists of a bottom-mounted acoustic current profiler, dozens of small selfcontained temperature sensors spaced every meter on the mooring line, and a few temperature/salinity sensors strategically placed within the different water mass layers expected at the site. In their last deployment, the moorings were positioned every 4 km on a cross-shelf line to observe the upwelling/downwelling fronts associated with the wind-driven evolution of a highly-stratified twolayered system on the inner continental shelf (Fig. 12). In this example, the initially flat isotherms were observed to form a strong downwelling front that Fig. 11. CalPoly's REMUS AUV with bioluminescence sensors. Fig. 12. Two snapshots of the high-resolution temperature section available every 15 minutes from the relocatable mooring array illustrating the formation of a subsurface downwelling front.

7 rapidly moved offshore in response to a strong downwelling favorable windfield. IV. CONCLUSIONS The New Jersey Shelf Observing System provides a well-sampled ocean for scientists and engineers to study processes and develop new instruments in a collaborative environment. The system is research oriented, i.e., it is run by scientists primarily for scientists, yet the data still has a tremendous societal impact. The websites displaying the real-time NJSOS datasets receive over 60,000 hits per day during the busy summer months, most from the recreational public. This same web-based real-time datastream is highly valued by the scientists desiring to deploy their sensors in the right place at the right time. Rather than the traditional expeditionary approach of loading all the scientists onto one ship and sailing off into the unknown, real-time awareness of the ever changing state of the ocean enables scientists to choose when and where they want to send their sensors. Rather than reducing the number of research vessels required, we find that the observatory increases the need for multiple research vessels, since each scientist wants to go to a different place to study their specific processes of interest. Our core fleet of three vessels is often augmented by additional research boats required to support the diverse needs of other scientists attracted by the collaborative environment. At times, up to nine research vessels were simultaneously operated on a single day within the observatory. For the last decade, operators of coastal observatories have maintained their systems by piecing together numerous small research grants. The inefficiency of this approach has been stated and restated in paper after paper and presentation after presentation. All agree that the only way to effectively operate coastal observatories is to develop a new source of sustained funding well above the level of the typical research grant. One solution to the funding issue has been the rapidly growing popularity of dedicated congressional appropriations. Even these, however, are sufficient to solve only the start-up problem, leaving the question of sustainability to the operators. The recent report from Ocean.US on a sustained and integrated ocean observing system ensures us that help is on the way. They estimate that the funds available for the U.S. observatory effort will grow to $500 million per year in new money by the end of the decade. This is comparable in magnitude to the total annual federal research budget for oceanographic science. This significant increase in new money must be paralleled by a similar increase in the recruiting and training of new people for the overall effort to succeed. The developing call is for new education programs in operational oceanography [19] that more closely parallel the meteorological paradigm. In meteorology, some atmospheric science students are steered into Ph.D. research programs, while others find equally rewarding careers at the bachelors and masters level working in operational meteorology. While operational oceanography is already a significant part of the Navy and NOAA, the educational programs fill existing needs not specifically tied to the development and operation of a national observatory. A new generation of bachelors and masters level oceanographers with hands-on training in an operational observatory is still required. Acknowledgements The New Jersey Shelf Observing System is supported by ONR, NOPP, NSF, NOAA/NURP and the Great State of New Jersey. This paper highlights research conducted with several academic, government and industry partners, including SeaSpace, Inc., CODAR Ocean Sensors, Webb Research Corporation, the Naval Research Lab (Washington D.C. and Stennis Space Center), California Polytechnical University, Mote Marine Lab, and Florida Environmental Research Institute. NJSOS is operated by members of the Rutgers University (R.U.) Coastal Ocean Observation Lab (COOL), including Trisha Bergmann, Louis Bowers, Bob Chant, Liz Creed, Mike Crowley, Rich Dunk, Chip Haldemann, John Kerfoot, Josh Kohut, Chhaya Mudgal, Matt Oliver, and Sage Lictenwalner. References [1] Glenn, S., O. Schofield, R. Chant, F. Grassle The New Jersey Shelf Observing System. Oceanology International Proceedings. [2] Schofield, O., T. Bergmann, J. Kohut, S. Glenn, A coastal ocean observatory for studying nearshore coastal processes. Backscatter, 12, [3] Schofield, T. Bergmann, P. Bissett, J. F. Grassle, D. Haidvogel, J. Kohut, M. Moline, S.M. Glenn, The Long-Term Ecosystem Observatory: an Integrated Coastal Observatory. IEEE Journal of Oceanic Engineering, 27: [4] Glenn, S.M., D.B. Haidvogel, O.M.E. Schofield, J.F.Grassle, C.J. von Alt, E.R. Levine and D.C.

8 Webb, Coastal predictive skill experiments at the LEO-15 national littoral laboratory, Sea Technology, April, pp [5] Glenn, S., T.D. Dickey, B. Parker and W. Boicourt, Long-term real-time coastal ocean observation networks. Oceanography, 13, [6] Glenn, S., W. Boicourt, B. Parker and T.D. Dickey, Operational observation networks for ports, a large estuary and an open shelf. Oceanography, 13, [7] Moline, M. A., Bissett, W. P., Glenn, S., Haidvogel, D., Schofield, O An operational multi-scale real-time long-term ecosystem observatory (LEO-15) for the coastal ocean. Ocean Optics XV 3: [8] Kohut, J., S.M. Glenn, and D. Barrick, Multiple HF-Radar system development for a regional Longterm Ecosystem Observatory in the New York Bight. American Meteorological Society: Fifth Symposium on Integrated Observing Systems, pp [9] Kohut, J., S. Glenn, D. Barrick A multisystem HF Radar array for the New Jersey Shelf Observing System (NJSOS). Oceanology International Proceedings. [10] Simonetti, P. and D. Webb A simplified approach to the prediction and optimization of performance of underwater gliders. Proceedings of the 10 th International Symposium on Unmanned Untethered Submersible Technology. [11] Creed, E.L., C. Mudgal. S.M. Glenn, O.M.E. Schofield, C.P. Jones and D.C. Webb, 2002, Using a fleet of Slocum battery gliders in a regional scale coastal ocean observatory, OCEANS 2002 MTS/IEEE, in press. [12] Robinson, A., and S. Glenn Adaptive sampling for ocean forecasting. Naval Research Review, 51, [13] Moline, M.A., R. Arnone, T. Bergmann, S. Glenn, M. Oliver, C. Orrico, O. Schofield, S. Tozzi Variability in spectral backscatter estimated from satellites and its relation to in situ measurements in optically complex coastal waters. Journal of International Remote Sensing. (In Press). [14] Tozzi, S., O. Schofield, T. Bergmann, M. Moline, R. Arnone Variability in measured and modeled remote sensing reflectance for coastal waters at LEO-15. Journal of International Remote Sensing. (In Press) [15] Bissett, W.P., O. Schofield, S. Glenn, J.J.Cullen, W.Miller, A. Pluddeman and C. Mobley, Resolving the impacts and feedbacks of ocean optics on upper ocean ecology, Oceanography,Vol. 14, No. 3, [16] Schofield, O., T. Bergmann, J. Grzymski, S. Glenn Spectral fluorescence and inherent optical properties during upwelling events off the coast of New Jersey. SPIE Ocean Optics XIV 3: [17] Bergmann, T., Schofield, O., Cullen, J., Glenn, S. Moline, M. A Concurrence of inherent optical properties and particulate organic carbon concentrations in the Middle Atlantic Bight: Applications of ocean color imagery in coastal waters. Ocean Optics XV 3: [18] Creed, E.L. and S.M. Glenn, Realtime data transmission between research vessel and shore command center. OCEANS 2000 MTS/IEEE Conference Proceedings, Vol. 2, pp [19] Mooers, N.K Operational Oceanography: Shall We Dance? EOS Transactions, American Geophysical Union, 81:11: