Multibeam Bathymetry

Multibeam bathymetry data were acquired using a SeaBeam 2112 multibeam echosounder, which was part of the scientific equipment onboard the Ronald H. Brown. The 12-kilohertz (kHz) system uses up to 150 beams covering approximately 150 degrees (°) swath width at depths less than 1,000 m and 120° swath width for depths up to 5,000 m. The swath width decreases to approximately 90° for depths greater than 5,000 m.

The operation of the SeaBeam system is dependent upon the depth of the water —the deeper the water, the longer the pulse width, the slower the ping rate, and the narrower the swath width. The multibeam swath width in the deeper regions was between 10,000 and 12,000 m (5.3 to 6.5 nautical miles (nmi)). This multibeam swath coverage allowed the survey line spacing to be established at 6-nmi  intervals with an east-west azimuth. The pulse width ranged between 7 and 20 milliseconds (ms), with ping intervals varying between 8 and 20 seconds. The SeaBeam system recorded the raw data files to a Silicon Graphics Incorporated (SGI) O2 computer. As the raw data files were received, they were automatically opened and closed with a file size limit of 6 megabytes (MB). This small file size enabled adequate and efficient data management procedures during editing and quality control. The total file size for the processed 2003 SeaBeam bathymetric data varied between .5 gigabyte (GB)) and 1.2 GB.

Sound Velocity

Sound velocity corrections were required by the SeaBeam acquisition system to account for water column temperature and salinity variations. The cast data were derived from three separate sources. A deep-water conductivity, temperature, and density (CTD) cast was obtained before the start of SeaBeam data collection. A Seabird Electronics 911 CTD probe was used for obtaining CTD cast data. The maximum depth obtained with this instrument was 3,000 m. The Levitus tables (LEVITUS94: World Ocean Atlas) were used to supplement the cast data with velocity values for water depths between 3,000 and 9,000 m. Levitus tables represent the annual average water velocity structure for a 1°×1° area centered at -66.5° S. longitude and 19.5° E. latitude. This water velocity profile is in the form of discrete (depth- and velocity-dependent) points where the depth is in meters and the velocity in meters per second. Additional speed of sound data were recorded using a Sippican, Inc. Expendable Bathythermograph (XBT). Cast data from the XBT were limited to range of 760- to 1,365-m depth. A Seabird Electronics SBE-21 thermosalinograph (TSG) was constantly monitored for surface sound velocity values. Monitoring the TSG values enabled the scientific crew to determine when additional sound velocity data were required. Sound velocity values were entered in the acquisition system for real-time application with beam-forming and beam-steering capabilities. Variations in sound speed due to temperature and salinity differences within the water column affect the ray path and two way travel time of each formed beam from the SeaBeam transducer to the sea floor. The correct sound velocity variations are thus required to account for any "bending" of the ray path, because these variations and any bending of the ray path will affect the final depth measurement of each formed beam. No evidence of multibeam swath cupping or frowning was evident during data acquisition and processing.

Data Processing

The SeaBeam MB41 raw data files collected aboard the Ronald H. Brown were transferred from the SGI computer to a processing laptop via file transfer protocol (ftp). Once the transfer was complete, the raw file was converted using CARIS 5.3 HIPS & SIPS hydrographic and sonar data processing software. The processing team maintained the same processing procedures as employed by NOAA hydrographic field units. Once the data were converted, a digital terrain model (DTM) or dirty DTM (a DTM that has not been cleaned for “noisy” data) was generated for visual detection of artifacts and missed depths. The next step entailed reviewing and editing the data with CARIS Swath Editor, followed with CARIS Subset Editor. Both editing processes allowed the team to eliminate data artifacts or outliers.

After data editing, the weighted mean grid (or DTM) was regenerated with a grid resolution of 150 m. A surface model colorized by depth and grey toned was exported as a TIFF image. The 150-m weighted grid was also exported as an ASCII XYZ dataset that was converted to a gridded Universal Transverse Mercator (UTM) file, which can be read with the Interactive Visualization Systems (IVS3D) Fledermaus software. The CARIS and IVS3D Fledermaus software was used for visual data interpretation by the USGS geologists. The exported ASCII XYZ dataset was also imported into Esri ArcGIS to generate a 150-m grid surface geographic information system (GIS) layer.

Final data products included the 150-m-resolution weighted grid image files, provided in colorized by depth with sun illumination and gray tone surface model formats, and are included in the ArcGIS project file.

During the August 2003 survey the team noticed a discrepancy of about 100 m across the swath between data taken on this survey and data taken in February 2003 in the same area. After the survey was completed, we noticed that the system was running in apparent depth mode. This mode outputs true position, but depth is processed to reproduce depth measured by a vertical echosounder that assumes a uniform 1,500-meter-per-second (m/s) sound velocity through the water column. The data can be corrected to true depth by applying sound velocity correction in a vertical path; no ray bending is required because apparent depth mode preserves true position of soundings. L-3 Communications Corporation agreed to provide the USGS with the procedure used to generate an apparent depth from true depth and sound velocity profile to confirm the conversion from apparent depth back to true depth. The raw data included in this publication are the data as they were collected. The related correspondence is available as part of this report.