Mapping Intertidal Oysters

This describes the successful application of acoustic classification technology for mapping distribution and internal structure of intertidal oyster reefs. A QTC VIEW™ seabed classification system was deployed in the channels of the Winyah Bay, South Carolina, USA. The purpose was to collect acoustic data at two frequencies and demonstrate the capability of the system for the identification of oyster reefs.

Two site specific surveys were made in an area surrounding the pier at Oyster Landing. The surveyed water depth ranged from 0.8 m to 4 m. The survey was completed during the highest part of the tide and groundtruth was made at low tide by walking across the reef. The bathymetric data were gridded and overlaid on an airborne IR image. Raw complex waveform data at frequencies of 50 kHz and 200 kHz were logged. Objective unsupervised classification was completed using QTC IMPACT™ software and recently developed objective clustering package. Classification results were plotted as a classified vessel track. The data were also interpolated to generate a full coverage plot and class colours were identified using a similar acoustic class - similar colour scheme.

Figure 1. Interpolated 50kHz classification draped on a bathymetric model generated with the 200 kHz depth data.

Visualisation of the raw echoes showed distinct differences between individual frequencies and from oyster and non-oyster substrates. Objective classification resulted in the identification of 10 classes for the 50 kHz data (Figures 1 and 2) and 6 classes for the 200 kHz (Figure 3). The difference in total number of classes is driven by the ability of the 50 kHz to identify more soft sediment types and by the 200 kHz providing better results in differentiating oyster shell condition.

The classification results were plotted adjacent to a photomosaic (Figure 4) made from oblique angle digital photos of the oyster reef. The results indicated a very strong correlation between the classes and the presence or absence of oyster shell. It is possible to infer subclasses of oyster reef corresponding to living vs. non-living oyster but a more rigorous groundtruth is required for confirmation.

Quester Tangent would like to acknowledge the tremendous support of everybody at the Belle W. Baruch Institute for Marine and Coastal Research for supporting the fieldwork. In particular we thank Dave Bushek and Laura Schmidt for logistical efforts and arranging the associated workshop. In addition we would like to thank Loren Coen of South Carolina Department of Natural Resources for interesting discussion on the oysters of Winyah Bay.

Figure 2. Classification interpolated for the 50 kHz. The ten classes are coloured based on their location in Q-space. The green polygons at the edge of the figure represent the approximate boundary of the waterway digitized from the satellite image.

[ TOP ]

Figure 3. Classification interpolated for the 200 kHz. The ten classes are coloured based on their location in Q-space. The green polygons at the edge of the figure represent the approximate boundary of the waterway digitized from the satellite image.

[ TOP ]

Figure 4. Photomosaic of the patch reef near the pier presented with the interpolated 50 & 200 kHz classification data. A - Mud with shell, B - Living oyster, C - Living oyster, D - Sandy mud, E - Living oyster.

[ TOP ]

View the pdf of the entire report: Mapping Intertidal Oysters with QTC VIEW QTC Technical Report SR41-03