Lake Borgne, Louisiana Debris Mapping

Transcription

1 Lake Borgne, Louisiana Debris Mapping Abstract Gary R. Davis, Paul L. Donaldson, Walter Simmons, Rebecca Quintal Science Applications International Corporation 221 Third Street Newport, RI USA Under contract to the National Oceanic and Atmospheric Administration (NOAA) Office of Coast Survey the Marine Science and Technology Division (MSTD) of Science Applications International Corporation (SAIC) is performing a full bottom search survey of Lake Borgne, Louisiana. The primary purpose of the survey is debris mapping in the wake of the 2005 Hurricane Katrina. In water depths of approximately 4 to 20 feet, MSTD has deployed two shallow draft boats equipped with SAIC s ISS-2000 data acquisition and navigation system. One boat is equipped with a vertical beam echo sounder and a Klein 3000 side-scan sonar to obtain 100% bottom search. The second vessel is equipped with a GeoAcoustics GeoSwath system which is expected to obtain full bottom imagery to a swath similar to that of the Klein side scan sonar while providing a bathymetry swath of up to eight times the water depth. This paper presents the mobilization effort, status of the on-going survey, data processing, a preliminary evaluation of the two systems for detection of debris in shallow water, and a preliminary evaluation of the GeoSwath system for swath bathymetry in very shallow water. This paper also presents the data pipeline for both systems from acquisition to preliminary data products. This data pipeline includes horizontal and vertical error attribution of the GeoSwath bathymetry data, processing through CUBE, and generation of Bathymetric Attributed Grid (BAG), as well as S-57 feature file generation from within SAIC s SABER processing package. 1. Introduction On the morning of 29 August, 2005, Hurricane Katrina stuck the Louisiana and Mississippi gulf coasts. The foot storm surge and 120 mile per hour winds caused catastrophic damage along the Gulf coast as far as 120 miles from the storm s center. As of May 2006 it was estimated that Hurricane Katrina was responsible for the death of over 1800 people, over 700 remained missing, and over $100 billion in damage. The clean up and rebuilding effort continues today, 18 months later 1. When the storm surge subsided, it deposited tons of debris into the surrounding water bodies creating hazards to vessel navigation and causing adverse effects to the local commercial fishing industry. In September of 2006 NOAA started survey work of the waters to identify and locate marine debris. This data is being used to assist in the removal of the debris to re-establish safe navigation and commercial fishing and to update the nautical charts throughout the affected area 2. NOAA has contracted SAIC s MSTD to survey square miles of Lake Borgne, Louisiana (Figure 1). The survey specifications required 100% side scan coverage and single beam echo 1

2 sounder data of the four sheets covering Lake Borgne to the inshore limit of 4 feet 3. The statement of work encouraged the use of emerging technologies, such as interferometric sonar or autonomous vehicles. SAIC chose to deploy two vessels to complete the Lake Borgne surveys, one equipped with a GeoAcoustics GeoSwath Plus Interferometric sonar, and the other equipped with a Klein 3000 side scan sonar and an Odom vertical beam echo sounder. Figure 1 SAIC Survey Area of Lake Borgne Showing Layout of Sheets A, B, C, and D. 2. Mobilization Mobilization for the survey commenced in early December 2006, with the installation of a tide station at the remains of Martello Castle in the southwest area of Sheet D at the entrance to Violet Canal (Figure 2). The station comprises two Design Analysis H350XL digital bubbler gauges, a tide staff, and GOES communication radios provided by John Oswald & Associates. 2

3 Figure 2 Tide station at Martello Castle, LA SAIC is employing two vessels to conduct the survey: The Lacey Marie (Figure 3) - an evolutional local work boat design called the Lafitte Skiff which is well suited for the shallow waters of Lake Borgne. She is 42 feet long, flat bottom with a 2-foot draft, wooden hull, and diesel powered. The vessel is owned and operated by Campo Marine, Shell Beach, LA. Figure 3 Lacey Marie The Thomas Dowell (Figure 4) A 32 foot aluminum work boat with two 150hp outboard motors. The vessel is owned and operated by Lowe Engineering. 3

4 Figure 4 Thomas Dowell Both vessels are configured with SAIC s ISS-2000 real-time survey system (Figure 5 and Figure 6. This system consists of a high-end dual processor computer with the Windows XP operating system, which runs SAIC s ISS-2000 software. This software provides survey planning and control in addition to data acquisition and logging for single beam, multibeam and navigation data. The system includes: POS/MV 320 Position and Orientation System with a Trimble Probeacon Differential Receiver Trimble 4000 GPS Receiver with a Differential Receiver Notebook computer for maintaining daily navigation and operation logs Seabird Model SBE Conductivity, Temperature, Depth (CTD) profiler Uninterrupted power supplies (UPS) for protection of the entire system For side scan and bathymetry, the Lacey Marie is configured with a 250 khz GeoAcoustics GeoSwath Plus Interferometric Bathymetric system. Sonar Frequency Maximum Water Depth Maximum Swath Width Range Resolution Across Track Two Way Beam Width Transmit Pulse Length 150m Swath Width 300m Swath Width Table 1 GeoSwath Specifications 250kHz 100 meters 300 meters Up to 12 x depth 1.5cm 0.5 Azimuth 8µS to 1mS 10 swaths per second 5 swaths per second The GeoSwath transducers are bow mounted on the Lacey Marie. The installation includes a pair of transducers (35cm by 15cm by 6cm) mounted on a V bracket. The V bracket also houses the heave/pitch/roll sensor, or motion reference unit (MRU) of the POS/MV. 4

5 Figure 5 ISS-2000 and GS+ Systems Installed in the Lacey Marie The Thomas Dowell is configured with a bow mounted Klein 3000 side scan for side scan and an Odom Echotrac CV vertical beam echo sounder with a hull mounted 200kHz transducer. Figure 6 ISS-2000 System installed in the Thomas Dowell 5

6 During the last week of December, 2006 through the first week of January 2007, SAIC mobilized for the survey. The mobilization effort included: Setting up a field facility in Shell Beach, LA consisting of a three bedroom mobile home and a 20 foot ISO container as a data processing and office facility. Installed survey equipment and sensors on two vessels, the Lacey Marie and Thomas Dowell. Surveyed sensor offsets on both vessels Conducted a Sea Acceptance Test of the installed systems including a patch test for determination of roll pitch, and heading biases on the GeoAcoustics Interferometric System on the Lacey Marie, settlement and squat measurements for both vessels (Lacey Marie and Thomas Dowell), and verification of positioning accuracies. 3. Survey Status As of the end of February, approximately 33% of the survey had been completed (7051 survey kilometers of the planned kilometers, Figure 7). Figure 7 February 2007, Progress Sketch 6

7 4. Data Processing Data Processing is being performed both on site and in the Newport, RI, SAIC office using highend dual processor computers. Side scan sonar data are reviewed for targets and quality in Triton Imaging s Isis sonar software both at the field office and in the Newport office. The GeoSwath data are initially processed using the GeoSwath Plus (GS+) acquisition and processing software and then processed with SAIC s SABER (Survey analysis and Area Based EditoR) software. The single beam data are also processed with SABER. Preliminary Processing in the Field At the end of each survey day, both vessels copy all of the survey data to an external hard drive and deliver it to the Shell Beach Data Processing office (Figure 8). There the data are copied to a Network Attached Storage (NAS). Figure 8 SAIC s Shell Beach Data Processing and Field Office. The raw GeoSwath data are filtered to remove outliers and sound velocity corrections are applied using the GeoSwath Plus software. This results in geo-referenced bathymetry swath (CBF) and side scan swath amplitude (swamp) data files on a line-by-line basis. The CBF files are converted to multibeam Generic Sensor Format (GSF) files using data conversion routines built into SABER and the side scan swamp files are converted to extended Triton Format (XTF) using conversion tools built into the GeoSwath Plus software. The Klein 3000 side-scan data are collected in XTF format and the Odom single beam data are collected in GSF format. Both GSF and XTF formats are compatible with SABER which is used to complete the processing and analysis to the final deliverable products. A time window file listing the times of all valid online side scan data is created along with separate side scan file lists for generation of a track line and coverage mosaic in SABER. The 7

8 mosaics are reviewed using tools in SABER to verify swath coverage and to plan further survey lines to fill in any data holidays. The single beam bathymetry data are manually edited to remove invalid depths and to identify shoalest soundings on wrecks and obstructions. Both the GeoSwath multibeam bathymetry and Odom single beam data files are then used to create track lines and a preliminary sounding grid in SABER. Both of these products are reviewed for quality and coverage. On a daily basis the data are binned to a gridded depth layer, populating each bin with the average sounding in that bin while maintaining its true position and depth. Both main scheme and cross line binned grids are created and used for cross line analysis. Results of these analyses are reviewed to determine adequacy of data and sounding correctors. Twice a week, all of the raw and processed data files are copied to an external hard drive and shipped to the Newport Data Processing Center. Final Processing The GeoSwath multibeam bathymetry files (in GSF format) are reviewed for navigation quality, corrected for daily draft, squat and final verified tides within SABER. Once correctors are applied the horizontal and vertical errors for each data point are estimated using SAIC s error model for the GeoSwath sonar. Developing this error model has been a large undertaking for this project and has been a joint effort between SAIC and GeoAcoustics. At the writing of this paper, SAIC are in the final stages of completing the final error model. Once the errors are estimated, the data are processed using the CUBE algorithms within SABER. A PFM grid with CUBE surfaces is generated and reviewed for data quality, edited as necessary, including setting designated soundings. Once a sheet is complete and the final CUBE surface is generated and reviewed, the PFM grid is converted to a Bathymetric Attributed Grid (BAG) for delivery to NOAA. The singlebeam navigation data are reviewed for quality, the singlebeam files are corrected for daily draft, delayed heave (POS/MV True Heave), and final tides in post-processing, while settlement and squat corrections are applied in real time. The final deliverable for the singlebeam data is as soundings in the S-57 feature file as opposed to a gridded format. The S- 57 file is generated in the SABER software and quality controlled using dkart Inspector. During examination of side scan sonar data, a side scan review log is generated. This review log contains information about each file, including the line begin and line end times, survey line name, corresponding multibeam, single-beam, and side-scan file names, line azimuth, holiday information, and notes pertaining to objects posing hazards to navigation (i.e. wrecks, obstructions, debris), and other points of interest (e.g. navigation aides). Other pertinent information regarding the interpretation of the imagery is also logged in the spreadsheet. A quality review of each side scan file is conducted using Triton Isis to replay the data. During this review the processor assesses the quality of the data and defines holidays in the data where the quality is insufficient to determine the presence of contacts. The times of these data holidays are entered into the side scan review log. Data holidays are generally characterized by: 8

9 Surface noise (vessel wakes, sea Acoustic noise clutter, and/or waves) Large, dense schools of fish Towfish motion (yaw and heave) Density layers (refraction) Electrical noise Turbidity clouds During side scan review, sonar contacts are selected and measured using the ISIS Target utility. Significant side scan contacts are chosen based on size and height or a unique sonar signature. In general, contacts with a height greater than or equal to 50 centimeters are selected. Wrecks and large objects are positioned at their highest point. Similarly, contacts for debris fields are positioned at the highest object in the debris field. Additional contacts are made on other man-made objects such as exposed cables, pipelines, and sewer outfalls. Additional information regarding objects not included as contacts but still noted in the side scan review log include descriptions of other man-made objects such as bottom fishing gear and non-significant objects. Weekly deliveries are made to NOAA which include TIFF images of all contacts made the previous week, an Excel document listing information about each contact (position, measurements, least depth, etc.) and a boundary file (.dxf) of the cumulative area surveyed to date. The final side scan deliverables will include a mosaic for each sheet in addition to the contact information and contact images. 5. Summary At the time of this writing it is not possible to draw strong conclusions about the effectiveness of the systems deployed in comparison to other systems. Progress on the surveys has been as expected and the data are of good quality. Several dangers to navigation have been reported to NOAA and have been applied to the charts covering Lake Borgne. SAIC expects to be able to present more definitive results from the surveys at the May 2007 conference. Based on preliminary analysis of the GeoAcoustics Interferometric data: Significant contacts with a height of 1 meter can be detected in the side scan imagery. Usable multibeam bathymetry swath is approximately 150 (75 to either side of nadir) or 7.5 times the depth below the transducer. 6. References 1) Hurricane Katrina-Most Destructive Hurricane Ever to Strike the U.S., NOAA Website 2) Gulf of Mexico Marine Debris Project, NOAA Website 3) NOAA Hydrographic Specifications and Deliverables, June 2006, 9