Navigating the Kill
October 27, 2011
All along the eastern seaboard of North America, seaports are preparing for a new wave of shipping traffic that is anticipated with the completion of the $5.25 billion Panama Canal expansion project in 2014. Although the increased economic activity will be welcome, there is a problem of logistics. Only one East Coast port, the Port of Virginia in Norfolk, is currently able to handle the large “post-Panamax” ships, which require draft depths of 48 feet or more.
At the Port of New York and New Jersey, which receives container ships from the Far East as well as from the Atlantic and Gulf Coasts, the Caribbean, Africa and the Persian Gulf, channels within the harbor range in depth from 30 to 45 feet. A $1.6 billion project led by the U.S. Army Corps of Engineers to deepen some of these channels is under way. The project includes 18 dredging contracts and the construction of four marsh restoration projects, in part to mitigate the impact of the dredging. Among the channels being deepened is the Kill van Kull (translated: channel from the pass or ridge), a tidal strait that lies between Staten Island, N.Y., and Bayonne, N.J.
A project of this magnitude requires detailed, highly accurate mapping. In a unique partnership, the Corps worked closely with seniors from the New Jersey Institute of Technology (NJIT) Surveying Engineering Technology (SET) program to apply an innovative combination of GPS and imaging techniques to acquire data and map the Kill van Kull.
The Kill van Kull tidal strait connects Newark Bay and Upper New York Bay. It is approximately 3 miles long and 1,000 feet wide. To accommodate the passage of larger ships, the strait has required continued dredging and deepening over the years.
Dredging operations in the Ports of New York and New Jersey began in the late 1960s and early 1970s with the construction of the South Elizabeth Channel. The Appropriation Act of 1985 permitted dredging of the Kill van Kull channel to accommodate large cargo vessels to dock in the New York and New Jersey ports. By 1997, the Kill van Kull was deepened to 45 feet in soft soils and 47 feet in bedrock to meet the demands of vessels with drafts of 41 feet or more.
Meanwhile, the size of vessels has continued to increase. Although the current capacity for ships passing through the Panama Canal is 5,000 twenty-foot equivalent units (TEUs), ships as large as 12,000 to 14,000 TEUs will be able to navigate the region when the canal expansion is complete. To meet the demands of these much larger vessels, the Corps must deepen the Kill Van Kull channel approximately 52 feet into bedrock.
Traditional surveying methods used to map channels for dredging projects were crude and costly. For example, before the adoption of multibeam sounding systems, the Corps would have used a mechanical bar sweep to map the channel. With this method, an operator monitors a bar attached to the barge as it is lowered to the project depth (while accounting for tide) and pushed along the channel. If the bar struck an object, the operator yelled, “Strike!” The operator then raised the bar incrementally until the object was cleared. The procedure was extremely labor intensive and required a large crew to support barge and bar activities.
Technology advances allowed the development of offshore positioning by radio beacon resections and single-beam transducers to develop dredging charts. Post-processing of single-beam sounding data required corrections for sea state parameters. In a single-beam system, little is known about the seafloor in between the cross-sections. While spacing between the sections could be reduced to minimize potential outcrop misses, the risk of missing a high spot is too great on critical jobs, particularly with rock seafloors such as the Kill van Kull. The Corps would have had to use multiple transducer systems to provide a sweep of the channel bottom for a near 100 percent ensonification of the seafloor.
Offshore positioning also presented challenges. Prior to modern GPS (RTK) centimeter-level precision positioning, surveyors used DGPS of L1 with differential correction broadcast by the Coast Guard to meter-level accuracy. Earlier methods included the use of robotic total stations with an intersection method called a “polar fix.”<
For the Kill van Kull channel mapping project, the Corps/NJIT team had the opportunity to take advantage of more recent advances in positioning and sounding technology to facilitate a rapid return of high-precision hydrographic survey results.
Multibeam surveying, in particular, has become the most sophisticated method of deep-water hydrographic surveying, providing increased bottom ensonification, accuracy, precision and efficiency. The high measurement rate of dual-frequency phase-based GPS measurements fills in the measurement gap between the heave sensor measurements and the water level gauge measurements.
Furthermore, surveyors can achieve improved estimates on the effects of squat, settlement, dynamic draft and mitigate changes in static draft since the GPS antenna offsets are at fixed distances from the sonar acoustic center. By combining GPS and inertial measuring unit (IMU) observations, surveyors can directly account for the effects on the water depth correction.
The Corps/NJIT team applied state-of-the-art GPS real-time network corrections from the Leica SmartNet network of reference stations to survey the Kill van Kull. Currently the network consists of 21 GNSS reference stations located in New York, Pennsylvania and New Jersey. Signal corrections are available 24 hours a day, seven days a week, with broadcast correction service available to network-ready RTK receivers. The real time reference information is “seamless,” allowing users to receive consistent correction data even as they move to different locations throughout a project. Data from each of the reference stations is received over the Internet at a highly secured server facility in Boston, where it is processed using Leica SpiderNet network calculation software and made available to users in real-time when they log in or as custom RINEX data for later downloads.
The SmartNet network provides higher offshore positional accuracy compared to traditional RTK methods. Using observations from multiple reference stations, SmartNet continuously monitors the integrity of reference station data and systematic errors, including ionosphere, troposphere, satellite orbit errors and multipath. The network then generates correction values for each individual satellite tracked from each of the reference stations. The MAC correction method then allows the rover to decide the most appropriate corrections to be applied at that time. This allows rovers to move freely around the area without having to worry about which base or corrections they are receiving or what processing method would be best at the time.
On the Kill van Kull mapping project, the GPS-derived water levels provided previously unmeasured water levels at the survey platform and vastly improved vertical accuracy to hydrographic surveys over discrete tidal zoning. Since the phase center of the GPS antenna and the acoustic center of the multibeam sonar are at a fixed distance and move in unison with vessel loading, the NJIT team was able to eliminate settlement and squat, vertical uncertainties in static draft, and dynamic draft.
In the past, surveyors would have had to rely on discrete tidal zoning provided by NOAA’s Center for Operational Oceanographic Products and Services (CO-OPS) to chart datum. GPS-RTK water levels are ellipsoidal referenced elevations and are used to derive corrections relative to the tidal datum. GPS-RTK tide observations for the Kill van Kull survey are tied to several tide gauges (TG) along the project. Some of the gauges are manually set and observed, while others are continually monitored by the National Oceanic and Atmospheric Administration (NOAA).
Early tests revealed that the National Geodetic Survey geoid model of the Kill van Kull survey region is well defined. That allowed the survey team to use the latest published geoid model for the application of the chart datum as determined by NOAA and published by the NWLON (National Water Level Observation Network). Local water levels (identified as mean low-low water, or MLLW) were determined by comparing GPS-derived water levels to a control station, a primary tide gauge that was continually monitored over the entire epoch period to obtain the MLLW values of the local project area for the survey epoch. However, in areas where the geoid model was questionable, a local calibration method was used that compared GPS data to differential leveling. In particular, a novel approach that compared GPS-derived tidal data collected over extended periods of time from barges located offshore eliminated the use of the geoid entirely.
The TG at Caven Point was located only a short distance from the survey area and could have been used for the Kill van Kull mapping project, with the team applying time and amplitude differences to achieve accurate tidal modeling. However, the location of the Kill van Kull deepening project is at a distance where the tide is different from Caven Point. Due to the critical nature of this project and the enormous cost of material removal, even minimal differences in applying tidal zoning were not acceptable. Instead the team used VDatum digital elevation transformation software developed by NOAA for coastal regions to determine the datum of MLLW for the project area. In addition, time and amplitude differences were obtained from NOAA. This information helped determine the placement of tide boards used to verify and check the RTK tides throughout the project.
The VDatum model revealed a 0.2 foot difference (specifically, from 2.78-foot to 2.95-foot below NAVD 1988) in the MLLW tidal datum across the project sloping toward New York Harbor. Because the current deepening project has been ongoing for several years, a superseded datum of 3.4 feet below NAVD 1988 is being held until the completion of the overall 50-foot Deepening Project. This means that the channel is dredged deeper than the charted depth.
Of course, neither the Corps nor private surveyors have the authority to publish determined water levels datums for charting purposes. A hydrographic survey relies upon published water level information provided by NOAA since it is the government agency responsible for publishing water level datums. A hydrographic survey by an agency other than NOAA can only verify and report discrepancies in published values--similar to the way a land surveyor might locate and observe an NGS passive control point.
Imaging technology was the second component of the hydrographic survey. Using high-frequency sonar, the team gathered bathymetric images and water depth measurements.
Sonar relies on the principles of sound velocity as it travels through the different mediums. Under water, the speed of sound is the distance traveled during a unit of time by a sound wave propagating through an elastic medium. The actual velocity differs greatly depending on the properties of the water column through which it travels, much like the way compression waves in solids depend on compressibility and density of the medium. For instance, in dry air at 20 degrees C (68 degrees F), the speed of sound is 343.2 meters per second (1,126 ft/s). This is 1,236 kilometers per hour (768 mph), or about 1 mile in 5 seconds.
A critical component for water depth measurement is a highly accurate sound velocity estimate. The Corps/NJIT team determined velocity directly from two types of equipment: a sound velocity probe (SVP) and a conductivity temperature depth (CTD) sensor. The SVP is perhaps the more common method; however, today’s modern CTD sensors offer huge advantages. These devices can obtain information in addition to just sound velocity, which can be used in the multibeam parameter set up, specifically temperature and salinity. In addition, some manufacturers have added GPS positioning and Bluetooth connection to their affordable CTD sensors. Traditional bar check methods can only determine the average sound velocity at a given depth and therefore cannot efficiently provide a full model of the water column. However, the physical bar check is used mostly to calibrate acoustic systems.
The basic principle of imaging using multibeam sonar is based on range measurement, angle measurement, beamwidth, seafloor ensonification, footprint dimension, and acoustic resolution. The seafloor is ensonified by a transmitted beam, and the sounding time is received and processed. The range is estimated from the return sounding time. The true depth of the water column is corrected for biases in the transmitted and returned beam angle, orientation and sound velocity.
The mobile surveying system for the Kill van Kull project comprised several survey devices, including DGPS/RTK devices, a positioning and orientation device (POS/MV-IMU), and a sound velocity probe. Relative offsets had to be precalibrated to obtain high-fidelity vessel attitude parameters during the survey.
In the case of sound velocity calibration, measurements were repeated many times throughout the course of the survey to maintain the highest level of accuracy depth estimates based on the sound velocity curve (velocity versus depth). The sound velocity was modeled and recorded so it could be applied to the time range of the sonar signal.
Critical time synchronization (i.e., latency) among positioning, navigation and depth sensors were required for high-fidelity volume computations. Within Kill van Kull, the vessel speed was limited to around 5 knots, although speed varies when traveling with or against current. Ping rate was based on spatial resolution. The sonar pings about 13 or 14 times per second in 50 feet of water. The unit cannot send its next energy wave until the previous energy has been returned. Therefore, the time between pings is a function of water depth.
The combination of GPS and imaging allowed the team to quickly gather high-accuracy precision marine geospatial data along the Kill van Kull. The Corps is using the dredging charts to facilitate the 50-foot Deepening Project, which is on schedule for completion in 2012.
The NJIT SET students presented their findings and experience from the Kill van Kull hydrographic survey at the annual NSPS Student Competition in 2011.