Bottom tracking issues and recognition thereof using SHOAL-300 green laser beam in dense fields of Zostera Marina and Laminaria Sp.

Description

This thesis assesses the ability of the SHOALS-3000 bathymetric lidar to correctly track the bottom in areas of dense aquatic vegetation, and explores a method to recognize improper bottom tracking, as a result of the vegetation, without ground truthing data. To perform the analysis, SHOALS-3000 and overlapping ship-based acoustic data were collected near the shores of Bonaventure and Paspébiac on the Bay de Chaleur, Quebec, in June 2006. EM3002 multibeam bathymetry and water column data, and Knudsen 200 kHz singlebeam water column backscatter were utilized for the SHOALS-3000 bottom tracking assessment. To assure a high level of vertical accuracy and exclude the necessity for real time, or predicted tide values post-processed kinematic (PPK) antenna solutions were allocated to the multibeam transducer face, by which multibeam solutions were matched with ellipsoid-referenced lidar solutions. An algorithm was developed that characterizes green laser waveforms, but excludes those that returned from land, or too shallow or too deep water. The algorithm includes three techniques to adapt as thoroughly as possible to obscure waveform bottom returns from vegetation soundings. A by-product of the characterization method is a term that describes the water clarity. The relative assessment revealed that the SHOALS-3000, compared to the EM3002, could achieve the manufacturer’s and IHO Order 1 accuracy specifications in unvegetated areas. Once aquatic vegetation is present, the bottom tracking of the SHOALS-3000 degrades: manufacturer’s and IHO Order 1 accuracy specifications could not be met. The green laser beam tracks the vegetation, or fails to return a sounding. This last observation is most concerning, implying that lidar datagaps, aside from reducing the effective coverage, are not necessarily due to lidar extinction depths, but potentially due to marine life covered navigational hazards. Waveforms from vegetation soundings have a small bottom return height, when compared to typical unvegetated seabeds, and as no sediment was found with similar reflectance, vegetated seabeds could therefore easily be discriminated from unvegetated seabeds. Vegetation identified with ground truthing data showed good agreement with small bottom return height values once the characterized waveforms were spatially plotted. A ratio describing typical bottom return height values of vegetated and unvegetated seabeds, confirmed the unambiguous reflectance of the laser, despite some day-to-day differences. This unambiguous behavior was incorporated with lidar bathymetry slopes to validate lidar soundings and identify fields of aquatic vegetation. Finally, overlapping keel mounted side scan sonar backscatter provides a potential to identify habitat based vegetation such as Laminaria sp. and Zostera Marina species.